Tiros Satellite Profile Data - NASA · Tiros N Satellite Profile Data James R. Scoggins, William E. Carle, Keith Knight, Vance Moyer, and Nine-Min Cheng Texas AGM Uuivevsity College

" .

NASA Reference Publication 1070

A Comparative Analysis of Rawinsonde and Nimbus 6 and Tiros N Satellite Profile Data

James R. Scoggins, William E. Carle, Keith Knight, Vance Moyer, and Nine-Min Cheng

NASA IC RP 1070 i

c.1 I

, I.

JANUARY 198 1

https://ntrs.nasa.gov/search.jsp?R=19810007125 2020-07-18T01:40:08+00:00Z


James R. Scoggins, William E. Carle, Keith Knight, Vance Moyer, and Nine-Min Cheng Texas AGM Uuivevsity College Stutiou, Texus

National Aeronautics and Space Administration

Scientific and Technical Information Branch

1981

ACKNOWLEDGEMENTS

The au thors express the i r s incere apprec ia t ion t o D r . William

Smith and Mr. Harold Woolf of the Nat ional Environmental Satel l i te

Serv ice for p rovid ing the satel l i te d a t a u s e d i n t h i s s t u d y , D r . Arthur

Dodd of t h e Army Research Off ice for h i s in te res t and support whi le the

study was in p rogress , and D r . Gregory Wilson for performing the

i n i t i a l computations. Finally, the authors thank Mrs. Karen H o o d f o r

he r a s s i s t ance i n p repa r ing t he f i na l manusc r ip t .

This research w a s supported by the U. S. Army Research Office,

under Grant N o . DAAG 29-76-G-0078 t o t h e Department of Meteorology,

Texas A&M University.

This report is published with the permission of t h e U.S. Army

Research Off ice for use in connect ion with s tudies ut i l iz ing space

technology for weather-related programs i n p r o g r e s s i n t h e Atmospheric

Sciences Division, Space Sciences Laboratory, NASA, Marshall Space Flight

Center.

ii

TABLE OF CONTENTS

Page

ACKNO.DGEMENTS. . . . . . . . . . . . . . . . . . . . . . . . ii

TABLEOFCONTENTS ....................... iii

L I S T O F T A 2 3 L E S . . . . . . . . . . . . . . . . . . . . . . . . . V i

L I S T O F F I G U R E S ........................ v i i i

1.1 statement of the Problem . . . . . . . . . . . . . . . . . 1

1 . 2 Previous Studies . . . . . . . . . . . . . . . . . . . . . 1

1.3 O b j e c t i v e s . . . . . . . . . . . . . . . . . . . . . . . . 2

2 . DATA UTILIZED . . . . . . . . . . . . . . . . . . . . . . . 4

2.1 Satel l i te D a t a . . . . . . . . . . . . . . . . . . . . . . 4

2.2 Rawinsonde Surface D a t a . . . . . . . . . . . . . . . . 4

3. ARBAS AiiALYZED AND SYNOPTIC CONDITIONS . . . . . . . . . . 5

-

3.1 Axeas Analyzed . . . . . . . . . . . . . . . . . . . . . . 5

3.2 Syncptic C o n d i t i o n s . . . . . . . . . . . . . . . . . . . . 5

4. INTER-AREA ANALYSIS OF THE DISCREPANCIES BETWEEN RAWINSONDE

A i i NDIBUS-6 DATA . . . . . . . . . . . . . . . . . . . . . 10

4.1 A p p r o a c h . . . . . . . . . . . . . . . . . . . . . . . . . 10

4.2 A n a l y s i s of D i s c r e p a n c i e s B e t w e e n R a w i n s o n d e and NIMBUS-6

. . . . . . . . . . . . . . . . . . . . " Profile P a r a m e t e r s 10

4.2.1 Temperature . . . . . . . . . . . . . . . . . . . . . 11

4.2.2 D e w - p o i n t Temperature . . . . . . . . . . . . . . . . . 1 4

4.2.3 T h i c k n e s s . . . . . . . . . . . . . . . . . . . . . . 15

iii

TABLE OF CONTENTS (Continued)

4.2.4 Mixing Ratio . . . . . . . . . . . . . . . . . . . . . 4.2.5 Precipitable Water . . . . . . . . . . . . . . . . . . 4.2.6 Lapse R a t e of Temperature . . . . . . . . . . . . . . 4.2.7 S t a b i l i t y . . . . . . . . . . . . . . . . . . . . . .

4.3 Analysis of Discrepancies Between Rawinsonde and NIMBUS-6 - - Data on Constant-Pressure Surfaces . . . . . . . . . . . .

4.3.1 Analysis Procedure . . . . . . . . . . . . . . . . . . 4.3.2 Temperature-related Variables . . . . . . . . . . . . 4.3.3 Dew-point Temperature . . . . . . . . . . . . . . . . 4.3.4 Geopotential Height and Geostrophic Wind . . . . . . .

"

- - 5 . SYNOPTIC STRUCTURE REVEALED BY RAWINSONDE AND NIMBUS-6 DATA

5.1 Constant-pressure Charts . . . . . . . . . . . . . . . . . 5.1.1 Temperature . . . . . . . . . . . . . . . . . . . . . 5.1.2 Dew-point Temperature . . . . . . . . . . . . . . . .

5.2 Cross Sections . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Taperatme ' . . . . . . . . . . . . . . . . . . . . . 5.2.2 Moisture-related Variables . . . . . . . . . . . . . .

6 . DETERMINATION OF W I N D FROM NIMBLE-6 SATELLITE SOUNDING DATA

6 .1 Satel l i te-der ived Winds on Constant-Pressure Surfaces . . . 6.2 Satell i te-derived Surface Wind . . . . . . . . . . . . . . 6.3 Comparisons of S a t e l l i t e - and Rawinsonde-derived Kinematic

"

Parameters . . . . . . . . . . . . . . . . . . . . . . . .

Page

16

16

17

17

18

18

20

25

28

33

33

33

33

36

36

36

4 1

4 1

46

47

i v

......... "- .. .........

TABLE OF CONTENTS (Continued)

Page

7 . COMPARISONS BETWEEN SIMULTANEOUS TIROS-N AND RAWINSONDE

DATA FOR 2100 GMT . ON 1 0 APRIL 1979 . . . . . . . . . . . . 51

7.1 Analysis . of Discrepancies Between Rawinsonde TIROS-N

Profi le Parameters . . . . . . . . . . . . . . . . . . . . 51

7.1.1 Temperature . . . . . . . . . . . . . . . . . . . . . 51

7.1.2 Dew-point Temperature . . . . . . . . . . . . . . . . 54

7.1.3 Thickness . . . . . . . . . . . . . . . . . . . . . 56

7.1.4 Mixing Ratio . . . . . . . . . . . . . . . . . . . . 56

7.1.5 Precipi table Water . . . . . . . . . . . . . . . . . 56

. . . . . . . . . . . . . . 57

7.1.7 S t a b i l i t y . . . . . . . . . . . . . . . . . . . . . . 57

7.2 Analysis of Discrepancies Between Rawinsonde and TIROS-N

Data on Constant-Pressure Surfaces . . . . . . . . . . . . 57 "

7.2.1 Temperature . . . . . . . . . . . . . . . . . . . . . 57

7.2.2 D e w - p o i n t Temperature . . . . . . . . . . . . . . . . 59

7.2.3 Lapse Rate and Horizontal Gradient of Temperature . . 61 - 7.2.4 Geopotential Height . . . . . . . . . . . . . . . . . 61

7.2.5 Geostrophic Wind . . . . . . . . . . . . . . . . . . 63 - 8 . SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . 67

8.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . 67

8.2 Conclusions . . . . . . . . . . . . . . . . . . . . . . . 67

REFEmNCES . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

.

LIST OF TABLES

T a b l e

".

1

2

3

8

9

L i s t of areas chosen for ana lys i s . . . . . . . . . . . . Mean (z) and s tandard deviat ion ((5) , lumped f o r a l l leve ls repor ted for each s ta t ion and f o r a l l s t a t i o n s in each area, of t h e mean discrepancy ( z ) , t h e a b s o l u t e mean discrepancy ( 18 I ) , and the root-mean-square discrepancy (RMSD), in degrees Celsius, between Nimbus-6- and rawinsonde-derived - temperatures and dew p o i n t s [6 S

(Tsat Trws)]. . . . . . . . . . . . . . . . . . . . . . Means and standard deviations of discrepancies (SAT-RWS) between s a t e l l i t e and weighted rawinsonde d a t a f o r selected parameters, by layer , synopt ic s i tuat ion, and geographical area . . . . . . . . . . . . . . . . . . . .

-

Means and s tandard deviat ions of normalized discrep- anc ie s i n t h i ckness (m km-') f o r t h e l a y e r s s u r f a c e t o 500 mb, 500 t o 300 mb, and 300 t o 100 mb f o r Areas I - I V . Means and standard deviations of discrepancies between gridded satel l i te and weighted rawinsonde parameters on selected constant pressure surfaces fox four geographical a r e a s . . , . . . . . . . . . . . . . . . . . . . . . . .

Average d i f f e rences and standard deviations of the d i f f e rences between sa te l l i t e -der ived (S) and hourly- observed (0) surface winds (S-0) €or t h ree r eg ions . . . Mean ( z ) and s tandard deviat ion ( 0 ) , lumped f o r a l l levels reported €or each s ta t ion and f o r a l l s t a t i o n s i n t h e SESAME region, of t h e mean discrepancy (61, t h e absolu te mean discrepancy ( 18 I ) , and the root-mean-square discrepancy (RMSD), in degrees Celsius, between TIROS-N and rawinsonde temperatures and dew po in t s [ a f (T - TR)] . . . . . . . . . . . . . . . . . . . . . . . . . . . Means and standard deviat ions of d i screpancies (S-R) between TIROS-N and rawinsonde da ta fo r s e l ec t ed parameters by layer f o r t h e SESAME region a t 2100 GMT on 10 A p r i l 1979 . . . . . . . . . . . . . . . . . . . . . .

S

Means and standard deviations of normalized discrepancies i n thickness €or the layers 1000 to 500 m b ( A ) , 500 t o 300 mb (B), and 300 t o 100 mb (C) f o r t h e AVE-SESAME area (m m-1) . . . . . . . . . . . . . . . . . . . . . .

Page

5

1 2

13

16

2 1

47

52

53

56

v i

LIST OF TABLES (Continued)

T a b l e Page

10 Discrepancies i n the Showalter Index derived from TIROS-N and rawinsonde data for the AVE-SESAME area . . . 58

11 Discrepancies in the Vertical Totals Index derived from TIROS-N and rawinsonde data for t he AVE-SESAME area . . . . . . . . . . . . . . . . . . . . . . . . . . 58

12 Means and standard deviations of discrepancies between gridded satel l i te and rawinsonde parameters on se lec ted constant pressure sur faces for the AVE-SESAME area a t 2100 GMT on 10 A p r i l 1979 . . . . . . . . . . . . . . . . 60

v i i

LIST OF FIGURES

Figure

1

2

3

4

5

6

7

8

9

10

11

Dis t r ibu t ion of rawinsonde (RWS) and Nimbus-6 soundings f o r A r e a s I - I V . , , . , , , . , . , . , , , . , , . , . Surface map covering Areas I, 11, and I11 a t 1800 GMT on 25 August 1975 ( c o n t o u r s i n m i l l i b a r s w i t h f i r s t one or t w o d ig i t s omi t t ed ) . . . . . . . . . . . . . . . . . Surface map covering Area IV a t 0600 GMT on 3 September 1975 ( con tour s i n mi l l i ba r s w i th f i r s t two d i g i t s omitted) . . . . . . . . . . . . . . . . . . . . . . . . Surface map covering Area V I 1 1 a t 2100 GMT on 10 Apri l 1979 ( con tour s i n mi l l i ba r s w i th f i r s t one o r two d i g i t s om*tted) . . . . . . . . . . . . . . . . . . . . Cumulative frequency distributions of discrepancies between sa te l l i t e and rawinsonde temperatures by layer f o r A r e a 1 . . . . . . . . . . . . . . . . . . . . . . . Locations of grid points and cross-sections €or four geographic regions . . . . . . . . . . . . . . . . . . . P r o f i l e s of the average difference and standard devia- t i o n of t he d i f f e rences between sa te l l i t e and rawinsonde t empera tu res ( s a t e l l i t e minus rawinsonde) f o r Areas I- IV...............,...........

P r o f i l e s of the average difference and standard devia- t i o n of the d i f fe rences be tween sa te l l i t e and rawinsonde v e r t i c a l lapse r a t e s of t e m p e r a t u r e ( s a t e l l i t e minus rawinsonde) for Areas I - I V . . . . . . . . . . . . . . . Prof i l e s of the average difference and standard devia- t i o n of the d i f fe rences between s a t e l l i t e and rawinsonde ho r i zon ta l t empera tu re g rad ien t s ( s a t e l l i t e minus rawinsonde) for Areas I-IV . . . . . . . . . . . . . . . Prof i l e s of the average difference and standard devia- t i o n of the d i f fe rences between s a t e l l i t e and rawinsonde dew-point temperatures (satellite minus rawinsonde) for Areas I-IV . . . . . . . . . . . . . . . . . . . . . . . P r o f i l e s of the average difference and standard devia- t i o n of the d i f fe rences between s a t e l l i t e and rawinsonde geopo ten t i a l he igh t s ( s a t e l l i t e minus rawinsonde) f o r

Page

6

7

8

9

1 4

19

23

24

26

27

Areas I-IV . . . . . . . . . . . . . . . . . . . . . . . 29

v i i i

Figure

LIST OF FIGURFS (Continued)

Page

1 2 p r o f i l e s of the average difference and standard devia- t i on o f t he d i f f e rences between satell i te and rawinsonde geostrophic wind speeds (satel l i te minus rawinsonde) f o r Areas I - I V . . . . . . . . . . . . . . . . . . 30

13 P ro f i l e s of the average difference and standard devia- t i o n of t h e d i f f e r e n c e s between satel l i te and rawinsonde geostrophic wind d i r ec t ions (satel l i te minus rawinsonde) for Areas I-IV . . . . . . . . . . . . . . . 31

1 4 Charts o f temperature and temperature difference (OC) a t 850 and 500 m b over the central United States region (Area 1) . . . . . . . . . . . . . . . . . . . . . 34

15 Charts of dew-point temperature and dew po in t d i f - ference ( O C ) a t 850 and 500 mb for the cen t ra l Uni ted States region (Area I) . . . . . . . . . . . . . . . . . 35

16 Cross sections of temperature and temperature difference ("C). fo r the cen t ra l Uni ted States region on 25 August 1975 a t 1700 GMT . . . . . . . . . . . . . . . . . . . . 37

1 7 Cross sections of dew-point temperature and dew poin t d i f fe rence ("C) f o r t h e c e n t r a l United S ta tes reg ion on 25 August 1975 a t 1700 GMT . . . . . . . . . . . . . . . 38

18 Cross sections of equivalent potential temperature and equivalent potent ia l temperature difference ("C) f o r t h e central United States region on 25 August 1975 a t 1700 G M T . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

19 P r o f i l e s of the average difference and standard devia- t i o n of the differences between satel l i te geostrophic wind speed computed from smoothed and unsmoothed he ights and rawinsonde wind speed f o r Areas I - I V . Differences were computed by subtracting rawinsonde from s a t e l l i t e values . . . . . . . . . . . . . . . . . . . . . . . . . 42

20 Profi les of the average difference and s tandard devia- t i o n of the d i f fe rences between satel l i te geostrophic wind d i r ec t ion computed from smoothed and unsmoothed heights and rawinsonde wind d i r ec t ion fo r Areas I-IV. Differences w e r e computed by subtracting rawinsonde from sa te l l i t e values . . . . . . . . . . . . . . . . . . 43

ix

LIST OF FIGURFS (Continued)

Figure

21

22

23

24

25

26

27

28

29

30

31

32

Page

Plotted winds and isotach analyses (m s ) a t 500 mb for t h e central United States region (Area I) . Isotachs were drawn from exact values and barbs p lo t ted t o t h e n e a r e s t 5 m s - 1 . ....................

-1

Plot ted surface wind and isotach analyses (m s ) f o r the cen t ra l Uni ted S ta tes reg ion (Area I) . . , . . , . . F i Ids of horizontal advect ion of temperature (10 OC s - ~ ) a t 850 mb for the cen t ra l Uni ted S ta tes reg ion (AreaI) . . . . . . . . . . . . . . . . . . . . . . . . . Pair ings of sa te l l i te sounding locations and rawinsonde s t a t i o n s a t 2100 GMT on 10 April 1979 . . . . . . . . . . Cumulative probabi l i ty f requency d i s t r ibu t ions o f temperature discrepancies within the layers 1000 t o 500 mb, 500 t o 300 mb, and 300 t o 100 mb f o r t h e AVE-SESAME area . . . . . . . . . . . . . . . . . . . . . . . . . . Cumulat ive probabi l i ty f requency dis t r ibut ions of dew-point temperature discrepancies w i t n i r , t h e l a y e r s 1000 t o 500 mb and 500 t o 300 mb f o r t h e AVE-SESAME area.

P ro f i l e s of average and s tandard deviat ion of d i f f e rences between s a t e l l i t e and rawinsonde temperatures ( s a t e l l i t e minus rawinsonde) f o r t h e AVE-SESAME area . , , . . , . , P r o f i l e s of average and standard deviation of differences between s a t e l l i t e and rawinsonde dew-point temperatures ( s a t e l l i t e minus rawinsonde) €or t h e AVE-SESAME area , ,

Prof i l e s of average and s tandard deviat ion of d i f f e rences between s a t e l l i t e and rawinsonde v e r t i c a l lapse r a t e s f o r t h e AVE-SESAME area . . . . . . . . . . . . . . . . . . . P r o f i l e s of average and s tandard deviat ion of d i f f e rences between sa te l l i t e and rawinsonde horizontal temperature g rad ien t s fo r t he AVE-SESAME area . . . . . . . . . . . . P r o f i l e s of average and standard deviation of differences between s a t e l l i t e and rawinsonde geopotential heights for f o r t h e AVE-SESAME area . . . . . . . . . . . . . . . . . Prof i l e s of average and standard deviation of di f fe rences between geostrophic winds computed from rawinsonde and sa t e l l i t e geopo ten t i a l he igh t s fo r t he Am-SESAME area. Differences were computed by subtracting rawinsonde from

-1

-6

s a t e l l i t e v a l u e s . . . . . . . . . . . . . . . . . . . . X

45

48

50

52

54

55

59

61

62

62

62

64

Figure

LIST OF FIGURES (Continued)

Page

33 Prof i l e s of average and s tandard deviat ion of d i f fe rence6 between rawinsonde winds and satel l i te-der ived geostrophic winds f o r t h e AVE-SESAME area. Differences were computed by subtracting rawinsonde from s a t e l l i t e v a l u e s . . . . . . 65

xi

A COMPARATIVE ANALYSIS OF RAWINSONDE AND NIMBUS-6

AND TIROS-N SATELLITE PROFILE DATA

James R. Scoggins, William E. Carle, Keith Knight, Vance Moyer, and Nine.-Min Cheng

Department of Meteorology, Texas A&M University

1 . INTRODUCTION

1.1 Statement of the Problem

Rawinsonde data have t radi t ional ly been the pr incipal source of

upper a i r atmospheric data. Recently, however, satell i tes have become a

major source of data and could allow improvement i n our knowledge of t h e

s t r u c t u r e of the atmosphere because: 1) sa te l l i t e soundings can be made

on a g loba l scale eliminating gaps in the data over the oceans; 2) a l l

measurements would be made by the same instrument so t h a t any e r r o r s

r e su l t i ng from t h e v a r i a b i l i t y between rawinsonde instruments would be

e l imina ted ; and 3) the sa te l l i t e measures t h e e n t i r e v e r t i c a l e x t e n t of

the sounding a t one t i m e so t h a t e r r o r s r e s u l t i n g from t h e downstream

d r i f t of the bal loon would be eliminated. However, b e f o r e t h i s new

source of data may be f u l l y u t i l i z e d , s t u d i e s must be done to de te rmine

t h e c a p a b i l i t i e s and l imi t a t ions of s a t e l l i t e d a t a f o r t h e p u r p o s e of

determining atmospheric structure.

1.2 Previous Studies

The f i r s t v e r t i c a l p r o f i l e s of both temperature and water vapor

w e r e determined from measurements of two infrared spectrometers carr ied by

the Nimbus-3 satel l i te . These da ta p rovided the f i r s t ana lys i s of t h e

three-dimensional thermodynamic s t ruc tu re of the atmosphere from sa te l l i t e

observations. The first s tud ie s (Wark and Hilleary, 1969; Hanel and Conrath,

1969) compared ind iv idua l sa te l l i t e temperature prof i les with corresponding

rawinsonde profiles; r e l a t i v e l y good agreement w a s found.

S t ae l in e t a l . (1973) found temperature d i f f e rences between Nimbus-5

and radiosonde profiles ranging between 1 and 4 K over an a l t i tude range

of 1 t o 20 km, with the largest d iscrepancies found a t the t ropopause and

near the surface. Layer-mean temperature d i f fe rences . . . between sa te l l i t e

* Research supported by U. S. Army Research Office, Research Triangle Park, North Carolina, under Grant DAAG 29-76-G-0078 t o t h e Department of Meteorology, Texas A&M University.

and radiosonde data for 13 pressure l e v e l s w e r e found by Waters e t al .

(1975) t o be 2.1 K i n December and 1.6 K i n June. Satel l i te-der ived

thicknesses were compared with rawinsonde layer thicknesses by Wilcox

and Sanders (1976). Standard deviations of 45, 49, and 115 m f o r t h e

l aye r s 1000-500, 500-250, and 250-50 mb, respectively, were found.

Kapela and Horn (1975) compared i s en t rop ic c ros s s ec t ions from 1200

GMT radiosonde data with those from Nimbus-5 soundings, and found agreement

with regard t o pat terns of isol ines , but considerably less d e t a i l i n t h e

sa te l l i t e cross sect ion than in the radiosonde cross sect ion. The same

w a s t r u e i n c r o s s s e c t i o n s of geostrophic and gradien t wind.

Smith " e t a l . (1975) used Nimbus-5 soundings to ob ta in geos t rophic

wind components perpendicular t o c ros s s ec t ions i n fou r separate case

studies. Their satell i te-derived geostrophic winds showed good corre-

spondence with observed winds as well as geostrophic winds derived from

radiosonde data. Arnold " e t a l . (1976) compared c ross s ec t ions of rawin-

sonde and Nimbus-5 temperatures and derived winds, and agreement w a s

found as t o g e n e r a l p a t t e r n s b u t s i g n i f i c a n t d i f f e r e n c e s i n c r o s s s e c t i o n s

of derived wind were present due t o d i f fe rences in hor izonta l t empera ture

gradients obtained from t h e two types of data . Horn e t a l . (1976) compared

cross sect ions of Nimbus-5 temperatures and derived winds from 1700 GMT

sa te l l i t e data with 1200 and 0000 GMT radiosonde data. They found the

sa te l l i t e pa t te rns to be cons is ten t wi th the changing synopt ic s i tua t ion ,

but with loss of d e t a i l .

"

In a study by Petersen and Horn (1977) , temperature prof i les obtained

from Nimbus-6 radiance measurements were used along with sea-level pressures

to cons t ruc t g r idded f i e lds of 500-mb geopotential height and geostrophic

wind over northeastern North America. Satel l i te-der ived winds obtained a t

1600 GMT were compared with geostrophic winds computed from 1200 and 0000

GMT rawinsonde height analyses. It w a s found t h a t t h e i s o t a c h f i e l d s of

geostrophic wind showed good con t inu i ty be tween s a t e l l i t e and bracketing

rawinsonde analyses. Locations of the 500-mb ve loc i ty maximums w e r e

reasonably consis tent between the two da ta sets. The r m s d i f fe rences

between satel l i te and rawinsonde geostrophic wind f ie lds ranged from

3.5 t o 5.0 m s . -1

2

Grody e t al. (1979) considered the use of microwave radiometric

measurements t o infer atmospheric wind f i e lds a s soc ia t ed w i th t rop ica l

storms. In an analysis of Nimbus-6 data through typhoon June in November

1979, s a t e l l i t e -de r ived winds were compared with 700-mb aircraf t

reconnaissance winds. Major d i f f e r e n c e s i n wind speed occurred primarily

near the storm center presumably because of t h e sa te l l i t e sensor ' s

insuf f ic ien t hor izonta l reso lu t ion .

1.3 Objectives

The primary objective of t h i s r e s e a r c h is the determination of how

w e l l quan t i t a t ive satel l i te data can be used t o dep ic t t he s t ruc tu re of

the atmosphere. This evaluation is made over a wide range of synoptic

and surface condi t ions by comparing Nimbus-6 and TIROS-N data with rawin-

sonde data in several geographic regions. Satell i te sounding data w i l l be

used to l oca t e f ron ta l zones and the tropopause, depict major features of

t h e wind f i e l d , and determine the distribution of temperature gradients,

moisture, and a i r mass s t a b i l i t y . Atmospheric structure determined from

satel l i te and rawinsonde data w i l l be compared.

3

2. DATA UTILIZED

2 .1 S a t e l l i t e D a t a

Satel l i te da t a u sed i n t h i s s tudy were provided by the.Nationa1

Environmental Sa te l l i t e Se rv ice . Nimbus-6 da ta inc lude temperature and

dew-point temperature a t 2 1 pressure l e v e l s (1000, 950, 920, - 850, 780, - 700,

670, 620, 570, - 500, 475, 430, 400, 350, 300, 250, 200, 150, 135, 115, and

- 100 mb) a t approximately 1700 GMT on 25 August 1975 and 0730 GMT on

3 September 1975. Nimbus-6 da ta fo r 1700 GMT on 5 February 1976 consist

of only 10 reported levels (underlined above) and t h e d a t a a r e of poorer

qual i ty than previous Nimbus-6 data because of deter iorat ion of t h e High

Resolution Infrared Radiation Sounder (HIRS) . TIROS-N data include

temperature and dew-point temperature a t ten p ressure l eve ls (under l ined

above) a t approximately 2100 GMT on 10 April 1979. A l s o included in the

Nimbus-6 and TIROS-N d a t a are the lat i tude, longitude, and the approximate

surface elevation for each sounding.

2 .2 Rawinsonde and Surface Data

""

Rawinsonde da ta €or use in comparisons with Nimbus-6 da ta were obtained

from the Texas A&M University archives of National Weather Service teletype

da ta , and from the National Climatic Center. Quantit ies used include the

temperature and dew-point temperature a t mandatory and s i g n i f i c a n t l e v e l s ,

and geopotential height and wind speed and direction a t mandatory l e v e l s

a t 1200 GMT on 25 August 1975, 0000 GMT on 26 August 1975, 0000 and 1200

GMT on 3 September 1975, 1200 GMT on 5 February 1976, and 0000 GMT on

6 February 1976. A s part of t he AVE-SESAME project, rawinsonde soundings

were taken a t 2100 GMT on 10 Apri l 1979. Twenty-one of these soundings

have been processed a t Texas A&M Universi ty for use in comparisons with

TIROS-N sounding data. Surface hourly data used in the study include

temperature, dew-point temperature, al t imeter sett ing, and wind speed and

d i r ec t ion a t 1700 GMT on 25 August 1975, 0700 GMT on 3 September 1979,

1700 GMT on 5 February 1976, and 2100 GMT on 1 0 A p r i l 1979.

4

3. AREAS ANALYZED AND SYNOPTIC CONDITIONS

3.1 Areas Analyzed

Eight geographical areas represent ing a wide range of surface and

synopt ic condi t ions were chosen for analysis . The da te ; time, and

locat ion of these areas are l i s t e d i n T a b l e 1. These areas represent

a v a r i e t y of sur face condi t ions inc luding f la t l and , mounta ins , and

water. Figure 1 shows the l oca t ion of Areas I-IV and t h e d i s t r i b u t i o n ,

of'rawinsonde and satel l i te data for each of these areas.

Table 1. L i s t of areas chosen for analysis. . .

of S a t e l l i t e Pass S a t e l l i t e N a m e

Area .

I

I1

I11

I V

V

V I

V I 1

V I 1 1

1700 GMT,

1700 GMT,

1700 GMT,

0730 GMT,

1700 GMT,

1700 GMT,

1700 GMT,

2100 GMT,

25 August 1975

25 August 1975

25 August 1975

3 September 1975

5 February 1976

5 February 1976

5 February 1976

10 A p r i l 1979

Nimbus-6

Nimbus-6

Nimbus-6

Nimbus-6

Nimbus-6

Nimbus-6

Nimbus-6

TIROS-N

Central U . S .

Caribbean

Canada

Western U. S . Central U. S.

Caribbean

Canada

Central U.S.

3.2 Synoptic Conditions

The sur face m a p a t 1800 GMT on 25 August 1975 i s shown in F ig . 2 . A

cold f ront extends from t h e Hudson Bay southwestward through the central

United States. The occluded part of the co ld f ront assoc ia ted wi th a

deep cyclone w a s l oca t ed i n t he ea s t e rn par t of Area 111. The polar a i r

w a s separated from the tropical a i r by the cold front extending through

Area I, while Area I1 w a s covered ent i re ly by an mT a i r mass. Horizontal

g rad ien ts of pressure and temperature were large i n Area 111, moderate i n

Area I, and small i n Area 11.

Figure 3 shows the su r f ace map i n t h e v i c i n i t y of Area IV a t 0600 GMT

on 3 September 1975. The area w a s covered by a modified mP or CP a i r mass

which w a s dry. Most of Area I V w a s f r e e from convect ive ac t iv i ty wi th

only a few thunderstorms in Arizona and New Mexico. Horizontal gradients

of pressure and temperature were small i n t h i s area.

5

a. Area I

c. Area I11

b. Area I1

d, Area IV

Fig. 1. Dis t r ibu t ion of rawinsonde (RWS) and Nimbus-6 soundings for Areas I-IV.

6

Fig. 2. Surface map covering Areas I, 11, and I11 a t 1800 GKT on 25 August 1975 (contours i n m i l l i b a r s w i t h f i r s t one 011' t w o d i g i t s o m i t t e d ) .

7

- . . . .... . ..

Fig. 3. Surface map covering Area I V a t 0600 GMT on 3 September 1975 (contours in mi l l ibars wi th first t w o d i g i t s o m i t t e d ) .

Synoptic conditions a t 1700 GMT on 5 February 1976 (not shown) include

a high-pressure ce l l centered over the Atlant ic Ocean t o t h e east of

South Carolina and a s ta t ionary f ront extending from West Vi rg in i a i n a

southwestward d i r e c t i o n t o c e n t r a l Texas. There w e r e s t rong grad ien ts of

temperature and dew-point temperature across the f r o n t i n Area V ( cen t r a l

United States). Flow i n Area V I (Caribbean) was dominated by the high-

pressure c e l l and t h i s area had r e l a t i v e l y weak g rad ien t s of temperature

and pressure. A t t h i s time, the re was no low-pressure center in Canada as

was present on 25 August 1975, so tha t t he f l ow w a s genera l ly from t h e

northwest in Area V I I . Temperature gradients in Area V I 1 were intermediate

between those of Areas V and V I .

The sur face map f o r Area V I 1 1 a t 2100 GMT on 10 A p r i l 1979 is shown i n

Fig. 4. A t t h i s time, a low-pressure system w a s cen te red i n Colorado. A

surface cold front extended from the low across Colorado, New Mexico, and

Texas i n t o Mexico. A w a r m front extended through eastern Texas across

Louisiana and Florida. Temperature gradien ts were moderate i n most of t h e

8

!

area of interest . Thunderstorms were reported along and i n f r o n t of

the cold f ront and much of t h e area was experiencing showers.

Fig. 4. Surface map covering Area VI11 at 2100 GMT on 10 April 1979 (contours in mil l ibars with first one or two d ig i t s omi t t ed ) .

9

4. INTER-- ANALYSIS OF THE D1SCRF;PANCIES BETWEEN RAWINSOaDE AND

NIMBUS-6 DATA

4.1 Approach

The general approach t o the analysis of both the rawinsonde and

satel l i te da ta and the comparisons between the two da ta sets is as

fol lows. Satel l i te soundings were compared with the closest sounding

loca t ion by determining the best estimate of the rawinsonde sounding a t

the t ime and location of t he satel l i te sounding. This w a s done by a

l i n e a r i n t e r p o l a t i o n ( i n t ine) of the rawinsonde sounding using the two

observations on e i t h e r s i d e of t h e s a t e l l i t e sounding. The p lo t t ed

soundings and the results obtained by comparing s a t e l l i t e soundings with

raMinsonde soundings made a t standard release times i n d i c a t e d t h a t t h i s

w a s the best approach. D a t a f ron t he sa te l l i t e and average rawinsonde

soundings for selected constant-pressure surfaces then were placed onto

a g r id ob jec t ive ly by computer and selected parameters computed from t h e

gridded data. The g r idded f i e lds were t r e a t e d s t a t i s t i c a l l y or analyzed

and compared. In addition, coxparisons were made between selected

ve r t i ca l c ros s s ec t ions of rawinsonde and s a t e l l i t e d a t a .

4.2 Analysis of Discrepancies Between Rawinsonde and Nimbus-6 P r o f i l e - Parameters

For t h e pu-pose of coxzparison, rawinsonde soundings were paired with

t h e c l o s e s t sa te l l i te soundings. N o t a l l s a t e l l i t e d a t a were used since

t h e r e were more sa te l l i t e than rawinsonde soundings. Seven parameters

were considered in this study: temperature, dew-point temperature,

mixing ra t io , th ickness , lapse rate of teqperature , precipi table water ,

and stabil i ty. Discrepancies between sa te l l i t e and rawinsonde data for

a l l seven parameters w e r e computed by subtracting rawinsonde from s a t e l l i t e

values. Computations were made for each level (e .g . , temperature) , or

each layer (e.g. , thickness), for each sounding. Additionally, discrepancies

were s t r a t i f i e d i n t o three layers : 1000 t o 500 mb, 500 t o 300 mb, and

300 t o 100 ~. Cumulative probabili ty frequency dis-LLibutions of the discrepancies

were cornputed for each layer for temperature, dew-point temperature,

thickness , lapse r a t e of temperature, and mixing r a t io f o r t h e ensemble

of a l l paired points within each layer.

10

4.2.1 Temperature

T a b l e 2 shows t h e m e a n and standard deviation of the m e a n discrepancy,

t he abso lu t e m e a n discrepancy, and the root-mean-square discrepancy (FMSD)

between Ni-nbus-6 and rawinsonde temperatures for Areas I - V I I . The

statistics were obtained from the lumped d iscrepancies for a l l l e v e l s

reported for each s ta t ion and f o r a l l s t a t ions i n each area t o provide a

s ing le set of cr i ter ia by which t o judge the resu l t s of the comparisons.

The mem discrepancy in temperature has an average which ranges from

0.2 t o 1.5.OC and a s tandard deviat ion which ranges from 0.4 t o l.O°C. This

ind ica t e s t ha t Nimbus-6 temperatures may be ei ther higher or lower than

rawinsonde-observed temperatures, but each algebraic mean i s a small

pos i t i ve number when averaged through the vertical column from the su r f ace

t o 100 m33 and over the whole area. The mean RMSD ranges from 1.1 t o 3.2OC

with largest magnitude in Area V. This may be due to the degrada t ion of

t h e H I R S da ta or changes in the meteorological conditions.

The means and standard deviations of temgerature discrepancies for

the 1000 to 500-, 500 to 300-, and 300 tcj 103-mb layers are shown i n

Table 3 €or Areas I - V I I . Mean discrepancies may be e i the r pos i t i ve o r

negat ive in the lowest layer , bEt are genera l ly pos i t ive in the xciddle

layer and are p o s i t i v e i n t he up;?er l aye r i n a l l seven areas. This

ind ica tes tha t sa te l l i t e -der ived tempera tures become increasingly higher ,

in generzl, than rawinsonde observed temperatures as higher layers are

considered. Magnitudes of the standard deviation range from 0.8 t o 3.7OC

and a re gene ra l ly smallest over the water (Areas I1 and V I ) . Smallest

s tandard deviat ions for each area general ly are found i n the middle layer ,

with the largest value in the upper troposphere, i .e. , tropopause region.

S t ae l in ” e t a l . (1973) have shown similar r e s u l t s ? and Smith ” e t al. (1975)

have shown t h a t i n the troposphere the discrepancies between sa te l l i t e

and rawinssnde soundings were genera l ly small except in the tropopause

region between 300 and 100 mb. The i r r e su l t s are i n agreement with

those presented in th i s s tudy .

The cumulative frequency distributions of t he d i sc repanc ie s i n

temperature are presented in Fig. 5 f o r Area I. The d i s t r i b u t i o n s are

approximately normal (straight l ines) except near the extremes. The

small sample s i z e i s inadequate for defining extremes of t he d i s t r ibu t ions .

11

Table 2. Mean ( z ) and s tandard deviat ion (a), lumped f o r a l l l e v e l s repor ted for each s ta t ion and f o r a l l s t a t i o n s i n e a c h area, of t h e mean discrepancy (F), t he abso lu t e mean discrepancy ( 13 I ) , and t h e root-mean-square discrepancy (RMSD) , in degrees Celsius, between Nimbus-6- and - rawinsonde-derived temperatures and dew po in t s [S Z (T - sat T r w s ) 1

I1 - Central U. S.

x U 8/25/75

v3 - Central U. S.

2/5/76

X U

Caribbean 8/25/75

~

VI - Caribbean 2/5/76

X

XI? - Canada 8/25/75

X U

V I I ~ - Canada 2/5/76

X

U

IV2 - X Western U. S.

9/3/75

'Twenty-one

Temperature

b PI RMSD

0.3 1.6 2.0 0.7 0.5 0.6

1.5 2.4 3.2 1.0 0.9 1.3

0.2 0.9 1.1 0.4 0.2 0.2

0.6 2.0 2.3 0.6 0.7 0.7

-

0.2 3. .9 2.3 1.0 1.0 1.0

0.2 2.6 3.0 0.7 1.1 1.2

0.4 1.8 2.2 0.8 0.6 0.7

~

"

Dew-Point Temperature Station Pairs

b .RMD

2.9 6.0 7.3 3.8 2.2 2.5

21

2.3 7.2 6.1 3.6

18

2.8 5.4 3.2 2.0

9

6.7 8.3 10.5 4 . 0 5.1 1 5.8

9

" . ". - 1

-2.0 5.5 4.6 2.2

7

- t- 7

- 6.7 8.8 9.9 7.0 5.4 5.3

23

. ~ .. . .L- .

levels from 1000 to 100 m b for temperature, 15 levels from 1000 to 300 mb for dew point.

to 300 mb for dew point.

300 mb for dew point.

2Sixteen levels from 700 to 100 mb for temperature, 10 levels from 700

'Ten levels from 1000 to 100 mb for temperature,5 levels from 1000 to

12

I ! j -

Table 3. Means and standard deviations of discrepancies (SAT-RWS) between sa te l l i t e and weighted rawinsonde data for selected parameters, by layer, synoptic situation, and geographical area.

j

1 I

Central United States Canada I Hestern u. s. ' Caribbean

I 172 25 Aug 1975 I 172 05 Feb 1976 I 172 25 Aug 1975 I 172 05 Feb 1976 1 172 25 Aug 1975 1 172 05 Feb 1976 I 072 03 Sep 1975 1

'A = 1000-500 nb '9 = 500-300 nb 'C = 300-100 mb

5 . . 'Dew-point data missing above 300 mb. Mixlng ratio data missing above 300 mb.

P w

Cumulative frequency

Fig. 5. Cumulative frequency distributions of discrepancies hetween s a t e l l i t e and rawinsonde temperatures by l a y e r f o r Area I.

The tendency for the cumulat ive f requency dis t r ibut ions to be s t ra ight

l i n e s when p lo t ted on probabi l i ty paper sugges ts tha t the d i screpancies

between s a t e l l i t e and rawinsonde temperatures are d u e t o random er rors .

Cumulative frequency distributions for Areas 11-VI1 (not shown)

revea l that the discrepancies for temperature are near ly normal f o r a l l

areas and a l l l a y e r s ; t h a t f o r dew-point temperature the lines are not

as s t r a i g h t as for temperature but to a f i r s t approximation may be

cons ide red s t r a igh t ; t ha t fo r mix ing r a t io t he d i s t r ibu t ions t end t o be

normal i n t h e two lower layers (data were not tabulated €or the upper

layer because of the absence of data) except on the t a i l s of t he

d i s t r ibu t ions ; and tha t t he d i sc repanc ie s €or the lapse r a t e of temper-

a ture within the three layers may be considered normally distributed.

4.2.2 Dew-point Temperature

The Nimbus-6 soundings of dew-point temperature do not appear t o be

a s r e l i a b l e as those of temperature for any of the seven areas. Table 2

shows t h e mean discrepancies an2 mean RMS d i sc repanc ie s fo r t he ve r t i ca l

column 1000 t o 300 m b for the seven areas . The nean FWS d i s c r e p n c i e s

range between 6.6OC (Area 11) and 9.9OC (Area IV) i n t h e f i r s t f o u r

1 4

r- regions, and vary from 8.3 t o 16.5OC i n Areas V-VII . Considering only

those areas with good q u a l i t y H I E d a t a (Areas I-IV), t he g rea t e s t

disagreement is found for the western United States where the a i r had

an extremely low water vapor content.

Discrepancies in dew-point temperature were examined f o r t h e 1000

t o 500- and 500 t o 300-mb layers. Neans and standard deviations of t h e

discrepancies within the two l aye r s fo r a l l seven areas are shown i n

Table 3. Large biases (mean d i f f e r e n c e s ) e x i s t i n t h e satel l i te da ta

re la t ive to the rawinsonde da ta . With the exception of Area V I I , t h e

mean d i f fe rence is smaller i n t h e l o w e r layer than in the upper layer .

This may b e a t t r i b u t a b l e t o the higher moisture content in the lower

layer than in the upper layer where the da t a were cons iderably no is ie r

than in the lower layer. Magnitudes of the s tandard deviat ion range

from 5.2 t o 13.4OC and ind ica t e l a rge d i spe r s ions of the d i screpancies

for each layer .

4.2.3 Thickness "

Thickness was computed from the satel l i te and rawinsonde data

according to

where R is the gas cons tan t for d ry a i r , T* the mean v i r tua l t empera ture

in t he l aye r between pressures p and p , and g is the accelerat ion due 1 2 t o gravi ty . Here ?* is given by

-

where w i s the mean mixing r a t i o f o r t h e l a y e r as determined from skew T-

log p p l o t s of rawinsonde and s a t e l l i t e p r o f i l e d a t a .

Layer thickness discrepancies were s t r a t i f i ed i n to t h ree l aye r s , i . e . ,

1000 t o 500 mb, 500 t o 300 mb, and 300 t o 100 mb. The thickness discrepancies

were normalized t o u n i t s of rn km-l*because of the var iab le th ickness of t h e

layers. Means and s tandard deviat ions of normalized discrepancies in

thickness are presented in Table 4 f o r Areas I-IV. Mean d i f f e r e n c e s i n

normalized thickness are similar t o t h o s e f o r temperature presented in

Table 3. The best agreement between satel l i te and rawinsonde-derived

thicknesses , indicated by the s tandard devia t ion of t he d i f f e rences ,

occurs in the middle layer , and the poorest in the upper layer ( t ropopause

region) . The smallest discrepancies occurred over w a t e r (Area 11).

15

Table 4. Means and s tandard deviat ions of normalized discrepancies i n thickness (m krn") for the l aye r s su r f ace t o 500 mb, 500 t o 300 mb, and 300 t o 100 m b f o r Areas I-IV.

1 Area I

I A B C Mean -1T8 ly9 6.b

DeV St* . I 6 . 2 4 . 8 1 0 . 0

da ta I 169 124 140 NO. Of

4.2.4 Mixing Ratio

I

A r e a I1

-5T4 -0.4 871 073 -1y5 376 -0T3 1T9 lT5 A B C A B C A B C

Area Tv A r e s I11

3 . 3 2 . 8 4 . 6 8 .1 5 . 7 8 . 3 8 .9 7 .5 10.1

81 54 54 138 138 157 54 42 49

Mixing r a t i o v a l u e s were obtained from dew-point temperature da ta

p lo t t ed on skew T-log p diagrams for rawinsonde and sa te l l i t e soundings.

Mixing r a t i o d a t a were s t r a t i f i e d i n t o two layers: 1000 to.500 & a d

500 to 300 mb. The r e s u l t s of comparisons between sa t e l l i t e and rawinsonde-

derived mixing ratios are presented in T a b l e 3. The means and standard

deviat ions of the discrepancies in the lower layer are greater than those

i n t h e upper layer €or a l l areas. These r e s u l t s were due t o t h e lower

moisture content in the upper layer where the da ta were considerably

n o i s i e r than in the lower layer . Satel l i te-der ived mixing ra t ios had a

negat ive bias re la t ive to rawinsonde-derived values in the lower layers

of Areas 11, 111, and V I I .

4.2.5 Prec ip i tab le Water

Prec ip i tab le water w a s computed by use of the equation

where w i s the p rec ip i t ab le w a t e r and the o ther symbols are as before.

Prec ip i tab le water w a s computed by integrat ing the mixing r a t io p r o f i l e

from 1000 t o 300 mb. A mean RMS discrepancy between profile pairs f o r

Areas I - I V of only 0.23 cm w a s found. This is somewhat be t t e r t han t he

0.5 cm RMS found by Hillger and Von der Haar (1977), presumably because

of t he microwave channels on Nimbus-6.

16

Means and s tandard deviat ions of discrepancies i n precipi table water

f o r Areas I - V I 1 a r e shown i n Table 3 . The r e s u l t s for these areas show

that average precipitable water may be obtained from s a t e l l i t e d a t a w i t h

an accuracy of about 1 m o r l e s s which i s qui te acceptab le in most cases.

The means were negative only i n two areas. The s tandard devia t ions in

Areas I-IV w e r e qu i te cons is ten t wi th a value around 2.3 mm except for

Area IV (western United States) where the moisture content was low.

Results obtained for Areas V and V I a r e similar t o t h o s e found i n

Areas I - I V while those obtained for Area V I 1 a r e much smaller.

4.2.6 Lapse Rate o f Temperature

Lapse r a t e s computed from Nimbus-6 and rawinsonde data were normalized

t o units of OC h-'. Discrepancies i n lapse rate were s t r a t i f i e d i n t o

three l ayers .

S t a t i s t i c s f o r t h e d i f f e r e n c e s between s a t e l l i t e and rawinsonde lapse

r a t e d a t a a r e shown i n Table 3 f o r a l l seven areas . In Areas I-IV, biases

i n the d i f fe rences a re wi th in 0.3OC km except for Area I V where t h e b i a s

i s -0.7OC km-' in the lowest layer. This large discrepancy i s caused by

e r r o r s i n t h e s a t e l l i t e d a t a n e a r t h e ground over the mountains. The

smallest standard deviation occurred i n the middle layer of each of t he

f i r s t f ou r a r eas w i th t he l owes t va lue over water (Area 11). Normalized

resu l t s ob ta ined for Areas V - V I 1 a r e s i m i l a r t o t h o s e f o r Areas I-IV

except the smallest magnitude of the s tandard deviat ion did not consis tent ly

occur i n the middle layer. This is probably due t o t h e u s e of only ten

l eve l s of data in Areas V - V I I .

-1

4.2.7 S t a b i l i t y

Showalter and ve r t i ca l t o t a l s i ndexes were computed f o r e a c h s a t e l l i t e

and rawinsonde sounding. Discrepancies between s a t e l l i t e and rawinsonde-

derived indexes were computed by subtract ing rawinsonde f rom satel l i te

values. The average and standard deviation of the d i f fe rences i n each

s t ab i l i t y i ndex were then computed f o r Areas I - I V .

It was found t h a t a l l Showalter indexes computed from s a t e l l i t e

data were posi t ive. This i s not ful ly understood but may b e r e l a t e d t o

the temperature and moisture structure of t h e a r e a s s t u d i e d , o r t o t h e

i n a c c u r a c i e s i n s a t e l l i t e dew-point and ambient temperatures in the lower

17

troposphere. The average and standard deviation of t h e d i f f e r e n c e s i n

Showalter indexes are 0.3 and 3.6, 1.4 and 2.8, 0.7 and 2.9, and -1.1 and

3.8 f o r Areas I, 11, 111, and IV, respect ively.

Smaller percentage errors i n t h e mean discrepancies were found f o r

the vertical to ta l s index than for the Showalter index. The average and

s tandard deviat ion of t h e d i f f e r e n c e s i n t h e v e r t i c a l t o t a l s i n d e x are

-2.1 and 2.0, -1.1 and 0.5, 0.4 and 3.1, and -1.6 and 4.1 for Areas I,

11, 111, and IV, respec t ive ly . The ver t ica l to ta l s indexes ob ta ined from

sa te l l i t e d a t a d i f f e r from those obtained from rawinsonde data by less

than 5%. This good agreement between satel l i te and rawinsonde data again

r e f l e c t s t h e h i g h q u a l i t y of t h e s a t e l l i t e t e m p e r a t u r e d a t a .

4.3 Analysis - of Discrepancies Between Rawinsonde and Nimbus-6 Data on - ” Constant-Pressure Surfaces

4.3.1 Analysis Procedure

An objec t ive ana lys i s scheme developed by Barnes (1964) w a s used t o

interpolate rawinsonde and s a t e l l i t e d a t a t o a square grid of 324 poin ts

with a grid-point spacing of 158 km. The gridding procedure i s i t e r a t e d

four times and a scanning radius determines the maximum d i s t ance t ha t a

da ta po in t may influence the grid-point values. A nine-point smoothing

rout ine (Shuman , 1957) w a s appl ied to each gr idded f ie ld to reduce

e f f e c t s of spurious var ia t ions. The gridding procedure, when used with

the proper scanning radius and the Smoothing routine, produces fields of

da ta which are similar t o hand-analyzed charts. Locations of the grid

po in t s are shown i n Fig. 6 f o r t h e c e n t r a l and western United States,

Canada, and Caribbean areas.

After t h e g r i d w a s established, sounding data from t h e s u r f a c e t o

100 mb were placed on t h e g r i d f o r t h e p a r t i c u l a r area involved. Data

sets were created with gr idded surface f ie lds of e leva t ion , p ressure ,

temperature, and dew-point temperature, and fields of temperature and

dew-point temperature a t each of the 2 1 p re s su re l eve l s (10 i n Areas V - V I I )

above the surface. This w a s done for both rawinsonde and sa te l l i t e data .

~n aux i l i a ry da t a set w a s created for rawinsonde-observed geopotential

height, and observed u- and v-component wind da ta a t the ten mandatory

levels. The 1200 and 0000 GMT rawinsonde gridded values were in te rpola ted

t o determine values corresponding to the time of the satel l i te da ta , a t the

18

r

a. Central United States

c. Canada

b. Caribbean

d. Western United States

Fig. 6, Locations of g r id po in t s and cross-sections for four geographic regions.

19

r i s k of incurr ing errors because of fast-moving map fea tures .

Differences between s a t e l l i t e and rawinsonde values were computed by

subtract ing rawinsonde f rom satel l i te values a t the g r id po in t s . The

mean and s tandard deviat ion of the d i f fe rences were prepared for nine

constant-pressure surfaces (850, 700, 500, 400, 300, 250, 200, 150, and

100 mb). V e r t i c a l p r o f i l e s of t hese s t a t i s t i c s a r e p re sen ted fo r each

parameter. T a b l e 5 conta ins the means and s tandard deviat ions of d i f f e r -

ences for each parameter on the 700-, 500-, and 300-mb su r faces fo r

Areas I - V I I . An estimated average magnitude of each parameter is given

i n the t ab le for the respec t ive cons tan t -pressure sur face over the a rea .

In cases where la rge g rad ien ts i n the parameter were evident, two values

appear that represent average values over portions of t he a r ea . The

magnitudes of the parameters are included in order to provide some idea

of the magnitudes of the d i f fe rences compared to the parameter under

consideration.

4.3.2 Temperature-related Variables

P ro f i l e s of the average and standard deviation of the d i f fe rences

between rawinsonde and Nimbus-6 temperatures are shown i n Fig. 7 f o r

Areas I - I V . The magnitudes of the average and standard deviation of the

d i f f e rences a r e r e l a t ive ly small i n Area 11, bu t a r e l a rge w i th more

v e r t i c a l v a r i a t i o n i n Areas I and I V . Average values i n Area I11 a r e

l e s s t h a n 0.75OC except near the tropopause (250 mb), and magnitudes of

the s tandard deviat ion are intermediate between those of Area I1 and

those of Areas I and I V . The f l a t t h e n n a l f i e l d i n the Caribbean,

associated with the weak an t icyc lonic c i rcu la t ion and high tropopause,

c r ea t e s optimum conditions for accuracy i n t h e s a t e l l i t e sounding data.

Average differences tend to be largest near the tropopause i n each of

t h e f i r s t f o u r a r e a s . These r e s u l t s a r e similar t o t h e 2OC RMS discrepancy

fo r t he lower troposphere found by Waters " e t a l . (1975) f o r t h e NEMS

instrument carr ied by t h e Nimbus-5 s a t e l l i t e , and a r e i n close agreement

w i t h t he 1.6OC RMS fo r t he 1000-500 mb layer found by Wilcox and Sanders

(1976).

P ro f i l e s of the average and standard deviation of the d i f f e rences

f o r l a p s e r a t e of temperature for Areas I - I V a r e shown i n Fig. 8. These

show t h a t up to nea r 400 mb, the average and standard deviation of the

differences are less than or equal to approximately 0.5OC km . Due t o

t h e v e r t i c a l smoothing i n the satel l i te soundings, the change of v e r t i c a l

-1

20

1 Table 5. I.lsans and s tandard deviat ioi is of d i s c r a ; a : i e s b-ltr<--.er, qriddecl s a t e l l i t e and weighted rawinsonde parameters

on s e l e c t e d c 0 n s t a r . t p r e s s z c s u r f a c s s f o r r o w g e o ~ r a n h i c ~ l a r e a s . I

T Tanper s t u r e Mean

( " C ) Standard Deviat ion Agsrox. #agnitude

Dew-point .),lean TeqGerature Standard Deviation

(OC) Approx. Magfiitude

~~ ~~

Magnitude of I-Iorizontal Mcin Gradien t of Standard Dcviatio.?

~~

tu Tm_oerature AppL?rox. Yagnitade P ("C/lOOO km)

Geopotent ia l N ea n Height Standard Deviation (m) ~ppr-cx. !.lagnitude

Zonal . Wind >lex? Speed Standard Deviation

(m s-1) Approx. Xlagnitude

Meridional Year! Vizd Speed Standard DCViatiGn (El s-1) Approx. >!=.,-nitede

S c a l a r Wind Mean Speed Standard Deviation (a s-1) Approx. Magnitude

k2nd Mean Di rec t ion Standard Deviat ion

(3eg) Appox. Magnitude

No. of d a t a p o i n t s

*Di roc t io ; l h igh ly va r i ab le

[ t i I I

!

I

i 7-

"

"

"

L

CenLJa l Un i t ed S t a t e s . Caribbean Area I Area V

700 500 3 00 7 00 500 3 00 700 500 300

Area I1

-0.7 0.4 0.3 1.4 -0.2 2.0 0.3 0.5 1.1 1.5 1.0 1.8

8.5 -7 -33 -21/-2.3 -26/-15 -50/-44 -1/9 -17/-35 -42,'-35 0.5 0.7 0.4 2.5 2.9 5.0

1 .6 1 .7 -1.0 -2.5 14.0 .6.3 3.3 4.0 - 4.8 6.0" 6.2 4.4 4.9 5.4 7.5 7.5 -

-25,"I.l -33/-40 - -3.5 -0.3 G.2

6.0 6.2 7.6 5.3 6.5 6.0 6.2 6.5 7.4 0 .1 0.1 0.1 1.7 1.1 0.8 c.4 3.5 0.5

"L7.3 -0.2 -0.0 1.1 0.3 1.5

-3.7 -0.3 0.4

1 1 1 5/30 6/13 6 l5/3 15/4 17/3 1.1 1.5 1.1 6 . 1 19.5 36.5 4.5 3.6 5.1

-1.4 -1.3 -0.6 -2 .1 7 .1 14 .1

7.4 4.8 5.1 -4.6 -1.0 11.7 23.4 21.7 30.4 14.0 17.6 31.9

! 3150 5623 9570 I 31CO 5sc3 94c3 8.3 10.2 15.0 45.1 44.9 79.7

-0.6 -5.2 1 . 2 -0.4 1.1 -3.6 -3.1 -0.7 10.7 5.1 7.3 11.1 5.1 5.8 7.6 13.0 12.9 31.6 io 25 j1 45,'-5 14 20

28 I -2 .3 - 2 . 1 -3.6 2.5 1.3 -13.6 1 -0.2 -1.4 -3.5

4.9 5 . 2 1C.2 1 1 3 . 3 1 G . 1 51.2 6.3 8.3 11.0 6 2 3 / 5 35/5 I S 1 2 " 1 L

I

4;: 4:; 0.9 1 4;; 4.2 21.7

2.5 1.3 2.7 10.3 12.6 15.0 51.5 4.7 5.4 5.6

28/6 45/7 2 3 23

-14.4 -13.9 -24.:: 10.5 1.3 -3.3 -2.6 4.3 -11.3 35.9 32.0 4C.3

* * * 240 250 240 225 250 230 64.6 79.4 116 62.5 C8.3 49.5

90 58 137

Area VI

700 500 300

0.4 2.4 4.2 0.7 0.7 0.8 6.5 -13 -40

3.7 9.5 -

-20 -3 5 -

-0.6 -0.6 0.2 0.2 0.2 0.8 6.0 6.7 6.7

7.4 6.7 - !

-0.3, -0.6 -0.1 2.2 2.6 3 .5

2 4 4

10.1 33.4 80.6 10.3 11.7 19.2 3180 5880 9650

2.0 2.0 2.3 6.5 6.3 11.1

-5/12 -12/15 -21/26

I

3.7 -0.7 -0.7 5.3 5.8 10.8

-14/13 -2/16 -10/26

-2.9 0.8 1.6 4.6 5.3 10.9 10 1 2

2o I 12.9 -22.1 -8.1 89.4 73.7 52.2

140 180 200

83

Tenpezature Mean ("C) S tandard Devia t ion

Approx. >!agnitude

Dew-point Mean Tevpera ture S tandard Devia t icn

("C) Approx. f.!~..rjnitude

Lasse Rate of 9' e : ., Tenpera ture S tandard Devia t ion

.. _.A-

('C km-l) Approx. Xagnitude

Magnitude of kior izontal >llean Gradient of Stanciard 2cviat ion Temperature Approx. !.lagnitude (oc/1ooo km)

~~~ ~ ~~

Geopotent ia l ? lean Height Standard Devint ior . (m) Appo:.;. ?lagnitccie

Zonal Wind >!ear! Speed Standard Deviatior: (m s-1) Approx. Elagnitcde

Ner id iona l >lea-: Wind S3eed Standard Dcviarior (m s-1) AF2rox. Magnitrde

Scalar Wind Mean Speed Standard Deviatior (m s"-) Approx. MagnituZe

Wind Mean Direc t ion S tandard Devia t ior

(deg) Approx. Magnitude

NO. of d a t a p o i n t s

:I I

J

"

Cacada Western U. S. Area I11 Area VI1

7 00 500 300 700 500 300 700 500 3 00

W e a N

0.2 -0.4 -0.3 4.2 0.1 1 . 0 -2.0 -0.5 0.8 1.1 1.1 1.6

1 U -9 -35 -24 -33 -52 -2 -16 -4 5 2.4 1 . 5 1.1 1.3 2.3 c .9

-4.3 -3.1 - 1 . 7 3 . 7 - 3.6 7.5 -9.4 3.8 6.0 - 8.4 4.9 - 4.0 7.5 12.5

__.

-30 , -44

0 . 1 0.1 -3.6 0.7 3.7 0.c i - -0.4 -0.G 0.4 0.3 0.7 5 .5 6 .2 6 .5

- 0.6 0.5 1 .4 1.1 1 . 2 - 6.7 7.3 5 .8 6 .3 4 .0 -

-0.4 -0.8 -1.6

5 7 io 10 5 3.5 8 8 - 6.4 3 .3 4 .2 6.4 2.9 4 . 1 3 .2 3 .3 - -3.2 1.7 2.2 3.4 3.4

1.3 -3.4 -1O.C 31.2 41.6 27.2 17.0 24.6 27.5 23.2 23.5

-13.9 -25.7 -27.0 27.0 56.2 62.7

22-53 5253 8GC.0 3100 5860 9600

0.5 1.4 -1.7 -1.4 1 . 0 7.0 8.4 11.7 8 .0 12 .1

2 a l a -1.4 -2.0 -3.3 2.4 1 . 5 1.1 0.4 1.2 1.6

7.9 7.2 7.0 4.3 8 . 2 1C.3 2 .8 6.7 9.7 -5 -15;lS -25/25 2 4 3

"

"

0.3 1 .0 1.6

13 25 30 5.1 8.7 12.0 6 .8 10 .2 12 .7 7.8 6.6 9.7 0.3 0.1 2.6 4 . 1 2 .0 -1.9

3 9 l a

~ 1:;: 2:;; iiii I 6:;: 4;;: 3:;;

I

-2.8 33.3 36.8

O m 3

*

a7 100 -

F Y

a E m 7 111

PC

100

150

200

25 0

300

400

500

700

850

- 2 - 1 0 1 2 3 4 Temperature Difference ("C)

100

150

200

F Y 250 300

Y E 3 400 PC 500

700

850

Area I

J

- 1 0 1 2

Temperature Difference ("C)

c, Area I11

100

150

200

250

300

400

500

700

850

100

150

200

Q Y 250 300

f 400

2 500 700

850

8 Q)

-2 -1 0 1 2 .Temperature Difference ("C)

b. Area I1

-2 - 1 0 1 2 3 Temperature Difference ("C)

d, Area IV

Fig. 7, Profiles of the average difference and standard deviation of the differences between sa t e l l i t e and rawinsonde temperatures (satel- l i t e minus rawinsonde) for Areas I-IV.

23

8 Y

d 7 m m al

PI &

100

150

- -

200 - < 250

- 400

- 300

-

500 - 700 - 850 - A v g

. I -0.5 0 0.5 1.0

Lapse Rate Difference ("c km 1 -1

a. Area I

100

1%

200

B Y 250 0) 300 & 7 400

PI 500

700

850

p1

-1.0 -0.5 0 0.5 1.0 ,Lapse Rate Difference ( "C km )

-1

c . Area I11

100

- 200

- 150

-

v ii 250

5

- 300 - 400

- 1 700

- 500

-

Av9

i!

850 - I I

/ U

-"-I- -1-0-0.5 O 0.5 1.0

Lapse Rate Difference ("C km -1 )

b. Area I1

100

150

200

7 00

850

-0.5 0 0.5 1.0

Lapse Rate Difference ("C km-')

d. Area IV

Fig. 8. Profiles of the average difference and standard deviation of the differences between s a t e l l i t e and rawinsonde ver t ical lapse ra tes of temperature ( s a t e l l i t e minus rawinsonde) for Areas I- IV.

24

l a p s e r a t e of temperature in satel l i te data associated with the t ropopause

occurs over a deeper layer than the corresponding change in rawinsonde

da ta , so t h a t t h e s a t e l l i t e d a t a i n d i c a t e a decrease which begins a t a

lower l eve l t han t ha t i n rawinsonde data. Therefore, differences tend

t o be nega t ive ( sa te l l i t e va lues too low) below the tropopause i n each

area, while approaching zero and perhaps changing sign above the tropopause.

This t rend i s par t icu lar ly ev ident i n Area I11 where the s ign of the

average difference changes a t 250 mb, the approximate level of the

tropopause.

Ver t i ca l d i f f e rence p ro f i l e s fo r Areas I - I V fo r t he ho r i zon ta l

g rad ien t of temperature are shown i n Fig. 9. Average d i f f e rences a r e

small , less than 2OC (1000 km) i n a l l f ou r a r eas , wh i l e s t anda rd

devia t ions a re near 1.7OC (1000 km) i n Area 11 and a r e l a r g e r i n

Areas I, 111, and I V where values reach 5OC (1000 km) . This i s i n

direct association with the magnitudes of the horizontal temperature

grad ien ts i n these areas . Area 11 (Caribbean) contains only small

g rad ien t va lues , thus a l lowing the d i f fe rences there to be smal l ; the

po la r f ron t i n Areas I and I11 causes gradients and d i f f e rences t o be

somewhat l a rge r . Average d i f fe rences show t h a t t h e s a t e l l i t e v a l u e s

a re too smal l in Area 111, too la rge i n Area 11, and vary i n Areas I

and I V from too small near the surface to too large through the middle

and upper troposphere.

-1

-1

-1

Resul ts for Areas I - V I 1 for temperature-related var iables are

presented i n Table 5. The average and standard deviation of the d i f fe rences

i n temperature , lapse ra te of temperature, and hor izonta l g rad ien t of

temperature general ly are larger i n Areas V - V I 1 than i n t h e f i r s t f o u r

a reas .

4.3.3 Dew-point temperature

V e r t i c a l d i f f e r e n c e p r o f i l e s f o r dew-point temperature are shown in

Fig. 10 for Areas I - I V . The standard deviation of t he d i f f e rences i n

Areas I and I11 averages approximately 5OC i n the lower troposphere, while

values of near 3.5OC and 7.5OC a re t yp ica l fo r Areas I1 and IV, respect ively.

Differences in Area I V a r e somewhat la rger than those in the o ther th ree

areas with values increasing above 400 m b t o n e a r 10°C. In a l l a r e a s .

except Area 111, t h e s a t e l l i t e i n d i c a t e s t o o much moi s tu re r e l a t ive t o t he

25

loo ." 13u t ( 200

- 250

- Y

5 300 -

m

SI ; 400

-

. I - 2 0 2 4 6 8

Temperature Gradient Difference ("C (1000 h) -1) a. Area I

loo [ 15C

200

- 250

-

- 300

400 - 500 - 700 - 850 -

I " - 2 0 2 4 6

Temperature Gradient Difference

c. Area I11 ( "C (1000 h)-1)

100

150

200

250

$300

400

8 500 700

850

al

m II)

PI

111 - 4 - 2 0 2 4 6

Temperature Gradient D ' f f erence ("C (1000 Ian) -4 b. Area I1

100 P-

150 - 200

- -25@

-

fi!

Y 8

Y

300

- 3 400

-

500 - 700

~ I

-/ 850

-

Avg

- 2 0 2 4 6 8 Temperature Gradient Difference

("C (1000 km) -1) d. Area IV

Fig. 9. Profiles of the average difference and standard deviation of the differences between s a t e l l i t e and rawinsonde horizontal temperature gradients (satel l i te minus rawinsonde) for Areas I-IV.

26

7 I

-5 0 5 Dew-Point Temperature

Difference (OC).

C. Area I11

-5 5 10 Dew-Point Temperature

Difference (OC)

b. Area I1

100

150

200

250

Q) 300 rn

400

. 500

700

850

Y

Y & PI

-5 0 5 10 Dew-Point Temperature

Difference (OC)

a. Area I

loo F 150

2oo t - 250 fi! Y

3 400 PI 500 al &

700

850

-5 0 5 10 Dew-Point Temperature

Difference ( O C )

d. Area IV

Fig. 10. Profiles of the average difference and standard deviation of the differences between s a t e l l i t e and rawinsonde dew-point temperatures ( s a t e l l i t e minus rawinsonde) for Areas I- IV.

27

8, 1.11.. I",,"-.. . .."-...----."-

rawinsonde with posit ive average differences a t most levels. Results

ob ta ined for Areas V - V I I , shown i n T a b l e 5, i nd ica t e that Nimbus-6

dew-point temperatures on 5 February 1976 are of .poorer qua l i ty than

those obtained on 25 August 1975.

4.3.4 Geopotential Height - and Geostrophic Wind

Geopotent ia l he ight f ie lds were computed from gridded satel l i te da ta

by integrat ing the hydrostat ic equat ion f rom the surface upward. In t he

integrat ion process , mean vir tual temperature for each layer w a s defined

as t h e arithmetic average of the values a t the top and bottom of the layer .

Surface temperature and dew-point temperature were obtained from hourly

synoptic data. Surface pressure w a s based on the a l t ime te r s e t t i ng

repor ted in hour ly t e le type da ta and the he ight o f the s ta t ion as given

i n each sa te l l i t e sounding.

Sa te l l i t e -der ived he ights are compared t o h e i g h t s c a l c u l a t e d by t h e

National Weather Service and supplied a t mandatory l e v e l s i n t e l e t y p e

da ta . Ver t i ca l p ro f i l e s of the d i f fe rences in geopotent ia l he ight a re

presented in Fig. 11 f o r Areas I-IV. Area I1 (Caribbean) exhibits the

smallest d i f fe rences between s a t e l l i t e and rawinsonde values, with standard

deviat ions increasing from near 8 m a t 850 mb t o 1 6 m a t 100 mb. Standard

deviation values range from 1 2 t o 56 m i n Area I, from 18 t o 50 m i n

Area I V , and from 28 t o n e a r 60 m i n Area I11 with maximum values near

250 mb. Average d i f fe rences are lower than standard deviations in a l l

areas except Area 11.

Pro f i l e s of the d i f fe rences between geostrophic winds computed from

rawinsonde and sa te l l i t e geopotent ia l he ights a re p resented in F igs . 1 2

and 13 fo r t he s ca l a r wind speed and wind d i r ec t ion , r e spec t ive ly , fo r

Areas I - I V . Average d i f f e rences i n wind speed are small in the lower

layers , wi th values near 5 t o 7 m s near the tropopause. Standard

deviat ions of d i f fe rences i n wind speed tend t o i n c r e a s e w i t h a l t i t u d e ,

with values between 5 and 15 m s . Differences near the tropopause tend

t o be larger than those elsewhere i n a l l f o u r areas. With the exception

of Area 11, the standard deviation of d i rec t ion d i f fe rence (F ig . 13) tends

to average approximately 45', and peaks near the tropopause. Area I1

(Caribbean) is q u i t e d i f f e r e n t from t h e o t h e r c a s e s d u e t o t h e small

wind speeds in tha t reg ion .

-1

-1

28

ii Y

R

PI m 1 m

PC

100 r 150

200

250

300

400

500

700

850

-20 0 20 40 60

Geopotential Height Difference (gpm)

a. Area I

100

150

200

250

300

400

500

700

850

0 -

20 40 60 Geopotential Height Difference

(gpm 1 c. Area I11

100

- 250

- 200

- 150

-

0 300 1 2 400 k"

500

- 700

-

850 - U

-20 0 20 40 60 Geopotential Height Difference

( g w ) b. Area I1

100

200

- 150

-

-

8 250 -

Y

Q) 300 - LC 7 3 400

- 700

- 500

- PI i

111 -40 -20 0 20 40 60

Geopotential Height Difference (gpm)

d. Area IV

Fig. 11. P r o f i l e s of the average difference and shandard deviation of t he d i f fe rences between s a t e l l i t e and rawinsonde geopotential heights ( s a t e l l i t e minus rawinsonde) fo r Areas I-IV.

29

I

100

150

200

3 250 300

Y

7 tn

k .g 400

500

700

850

-10 0 10 20 30

Scalar Difference (m s )

a. Area I

-1

150

200

250

$ 300 8.400

500

700

850

Y

m m

PI

L

-10 0 10 20

Scalar Difference (m s ) -1

c. Area I11

100

150

200

8 250

8 300 4 400 500

700

850

Y

7

M al

Scalar Difference (m s ) -1

b. Area I1

200

9 250 Y

3 400

500

700

850 J -10 0 a 10 20 30

Scalar Difference (m s )

d. Area IV

-1

Fig. 12. Profiles of the average difference and standard deviation o f the differences between sa t e l l i t e and rawinsonde geostrophic wind speeds ( sa t e l l i t e minus rawinsonde) for Areas I-IV.

30

I

I

I ,

loo r 150

200

-250

300

i400

2500

700

850

Y 3

3

U

Wind Direction Difference (deg)

a. Area I

100

150

200

i250 Y

5300 :400

500

700

850

el P I '

-20 0 20 40 60 30 Wind Direction Difference (deg)

-c.. .':-ea 111

100 - 150 - 200 -.

Y $400 - al

2 5 0 0 -

850

700 I / I

3 I

- 20

Wind Direction Difference (deg)

b. Area I1

100

150

200

3.250 Y

: 300 400

500

700

850

01 h

-20 0 20 40 60 80 Wind Direction Difference (deg)

d. Area JX

Fig. 13. P r o f i l e s of the average difference and s tandard deviat ion of t h e d i f fe rences between s a t e l l i t e and rawinsonde geostrophic wind d h e c t i o n s ( s a t e l l i t e minus rawinsonde) f o r Areas I-IV.

31

Results obtained for geopotent ia l height and geostrophic wind f o r

Areas V-VI1 are presented in T a b l e 5 a long wi th those ob ta ined for the

f i r s t f o u r areas. A comparison of t h e r e s u l t s from t h e two sets of

a reas i nd ica t e t ha t t he qua l i t y of sa te l l i t eder ived geopotent ia ; he ight

and geostrophic wind speed a re poore r i n Areas V-VI1 than in Areas I-IV.

Large differences between s a t e l l i t e - and rawinsonde-derived f ields may be

due t o t h e poor qua l i t y of t h e s a t e l l i t e sounding data and the use of

t e n - l e v e l s a t e l l i t e d a t a i n Areas V-VII.

32

5. SYNOPTIC STRUCTURE REVEALED BY RAWINSONDE AND NIMBUS-6 DATA

Analyzed constant-pressure char ts and cross sec t ions are presented

f o r Area I (central Uni ted States on 25 August 1975) . These were constructed

from gridded data, and represent the horizontal and ver t ical var ia t ions of

atmospheric parameters as depicted by satel l i te and rawinsonde data, as w e l l

as va r i a t ions of quan t i t a t ive d i f f e rences between the two types of da ta .

5.1 Constant-pressure Charts

5.1.1 Temperature

F ie lds of temperature a t 850 and 500 mb f o r Area I are presented in

Fig. 14. There is a surface f ront across the northwest port ion of t h e

area (see Fig. 2 ) that corresponds t o the higher-than-average temperature

gradient which is apparent in both types of d a t a i n that p a r t of the area.

A t 850 mb, rawinsonde temperatures range fran near 16OC jus t sou th o f t he

f r o n t t o n e a r 8OC nor th of the f ront , whi le sa te l l i te temperatures range

from near 16OC south of t h e f r o n t t o n e a r 6OC nor th of the f ront . This

set of cha r t s shows that , whi le temperature differences are l a r g e s t j u s t

south of the f ront over Missouri a t 850 mb (near 4OC) , a reasonable

correspondence exists between sa te l l i t e and rawinsonde temperature data.

The sign of the d i f fe rences seems t o be r e l a t ed t o t he l oca t ion of the

f ron t , s ince pos i t i ve d i f f e rences are t o the nor th and negat ive d i f fe rences

t o the south of t he f ron t .

It has been determined from analyzed charts for Areas I - I V t h a t

measurements of temperature obtained from satel l i te-observed radiancies

a re accura te enough t o d e p i c t f r o n t s on constant-pressure charts, al though

the con t r a s t i n sa te l l i te temperatures across the f ront is less than that

i n rawinsonde temperatures. Temperature patterns on constant-pressure

cha r t s from rawinsonde and Nimbus-6 da ta are similar.


Charts showing f i e l d s of sa te l l i te and rawinsonde dew-point temperatures

f o r Area I a t 850 mb are shown in Fig. 15. The satel l i te da ta are cons is ten t

with the rawinsonde data in terms of the genera l pa t te rn , wi th ind ica t ions

of moist a i r south of t h e f r o n t and dry a i r nor th of t he f ron t . The

gradients in dew-point in the sa te l l i t e da ta a t t h i s l e v e l are not

su f f i c i en t t o p rov ide p rec i se i nd ica t ion of t he f ron ta l pos i t i on . On t h e

other hand, the f ront can be located fa i r ly easi ly in t h e rawinsonde data

33

1 L J

a. Rawinsonde (RW), 850 mb

I b. Sate l l i t e (S), 850 mb

C .

L A .

Differences (S-RW), 850 mb

Fig. 14. C h a r t s of temperature and

d. Rawinsonde (RW), 500 mb

e. S a t e l l i t e (S), 500 mb

f . Differences (S-RW) ,500 mb

temperature difference (OC) a t 850 and 500 mb over the cen t ra l Uni ted States region (Area I ) .

34

a. Rawinsonde (RW), 850 mb d. Rawinsonde (RW), 500 mb

b. Sate l l i te ( S ) , 850 mb e . Sa te l l i t e ( S ) , 500 mb

. Difference (S-RW), 500 mb

Fig. 1 5 . C h a r t s of dew-point temperature and dew point difference ("C) a t 850 and 500 m b for the central United States region (Area I ) .

35

Since the gradien t -is quite s t rong ' in a band - f r o m Michigan t o Colorado.

Differences (Fig. 15c) are general ly between O°C and 5OC, although values

of 10°C occur just behind the f ront . As with t empera ture , the f ront

marks a l i ne s epa ra t ing pos i t i ve d i f f e rences t o the no r th from negat ive

d i f fe rences t o the south.

The dew-point temperature map for rawinsonde data a t 500 m b i n Area I,

also presented i n Fig. 15, shows many areas of strong gradients of moisture

which are no t p re sen t i n t he sa te l l i t e data . Maximum di f fe rences are

located paral le l t o and jus t south of the f ront with values reaching

10°C. These differences do not correspond with cloudiness.

5.2 Cross Sections

The l ine of the c ross sec t ion for Area I is shown in F ig . 6. Each

f igu re of c ross sec t ions p resented conta ins th ree parts: 1) a cross

sect ion der ived from rawinsonde data; 2 ) a cross section derived from

sa te l l i t e data; and 3) a cross sec t ion of differences expressed as

s a t e l l i t e minus rawinsonde values.

5.2.1 Temperature

The cross sec t ion of temperature for Area I (Fig. 16) shows t h e f r o n t

in the nor thern part of t h e s e c t i o n t o b e r e l a t i v e l y weak i n terms of

temperature contrast in the rawinsonde data, and weaker i n the sa te l l i t e

data. This makes t h e f r o n t d i f f i c u l t t o l o c a t e i n t h e s a t e l l i t e c r o s s

Sect ion, but nei ther type of da ta loca tes the f ront except as being

somewhere i n a broad zone of baroclinity. The f r o n t w a s located by use

of rawinsonde soundings, and the frontal position obtained also w a s used

with the satel l i te data . One f ea tu re of the d i f fe rence c ross sec t ion is

the presence of negat ive differences through most of the t roposphere in

the a i r south of t h e f r o n t . A l aye r o f pos i t i ve d i f f e rences ( s a t e l l i t e

values too high) is p resen t j u s t under the t ropopause in both a i r masses.

These differences apparently are the resul t of v e r t i c a l smoothing.

5.2.2 Moisture-related Variables

The cross sec t ion of rawinsonde dew po in t fo r Area I (Fig. 1 7 ) shows

a mois ture increase across the f ront from nor th to south assoc ia ted wi th

p re f ron ta l shower a c t i v i t y , and f a i r ly s t rong con t r a s t ac ross t he f ron t .

The satel l i te sec t ion i nd ica t e s much less cont ras t across the f ront wi th

highly smoothed pat terns . Differences are l a r g e s t where the rawinsonde

36

I

I O 0

200

30 0

500

850

" " " -30

747 645 532 433 229 S t a t i o n I d e n t i f i e r s a. Rawinsonde (RW)

l O O r 1

2 00

' O 0 I 50 0 c - 850 - -

L I I I L +g 74 7 229 645 532 433

S t a t i o n I d e n t i f i e r s

100

20 0

300

500

850

"""""2

747 645 532 433 229 S t a t i o n I d e n t i f i e r s c. Difference (S-RW)

Fig. 16. Cross sections of temperature and temperature difference ("c) fo r t he cen t r a l Un i t ed States region on 25 August 1975 a t 1700 GMT. (See Fig. 6 fo r pa th of c ross sec t ions . )

37

I I I

300

500

850

74 7 645 229 S t a t i o n I d e n t i f i e r s a. Rawinsonde (RW)

rnb

200

300

850

I oc rnb

200

300

500

850

I ,\ I - L

74 7 645 532 433 229 S t a t i o n I d e n t i f i e r s b. S a t e l l i t e (S)

I

747 645 532 433 229 S t a t i o n I d e n t i f i e r s

c. Difference (S-RW)

Fig. 17. Cross sections of dew-point temperature and dew point d i f fe rence ("C) €or the central Uni ted States region on 25 August 1975 a t 1700 GMT. (See Fig. 6 for path of cross Sections.)

38

gradients are largest . Without-pr ior knowledge of the approximate location

of t he f ron t , it would b e d i f f i c u l t t o l o c a t e t h e f r o n t on e i t h e r t h e

s a t e l l i t e c r o s s s e c t i o n of temperature or dew-point temperatuye.

c r o s s sec t ions of equivalent potent ia l temperature for Area 1 a r e

shown in Fig. 18. The d i f f e r e n c e i n a i r mass s t ruc tu re and s t a b i l i t y i s

shown in both types of da ta for th i s a rea . This appears to be a r e l i a b l e

v a r i a b l e t o examine from sa t e l l i t e da t a fo r dep ic t ion of f r o n t a l c o n t r a s t s

between a i r masses. Differences i n equivalent potential temperature

measurements are largest near the surface where the l a rges t d i f f e rences i n

moisture measurement occur.

39

7 -

S t a t i o n I d e n t i f i e r s a. Rawinsonde (FW)

-60

747 645 532 433 229 S t a t i o n I d e n t i f i e r s b. S a t e l l i t e (SI

100 - mb

. .

200- ((- -

Fig. 18. Cross sections of equivalent potential temperature and equivalent potent ia l temperature difference ("C) f o r t he cen t r a l Un i t ed S t a t e s region on 25 August 1975 a t 1700 GMT. (See Fig. 6 for pa th of c ross sec t ions . )

40

I - 6. DETERMINATION OF W I N D FROM NIMBUS-6 SATELLITE SOUNDING DATA

Objective methods of computing upper-level and surface wind f i e l d s

f run Nimbus-6 satell i te thermodynamic da ta were developed. Satel l i te-

derived and rawinsonde wind f i e l d s ar.e compared on gridded constant-

p r e s s u r e c h a r t s a t n i n e pressure l e v e l s i n Areas I-IV. Rawinsonde winds

used in the compar isons a re l inear ly in te rpola ted to cor respond i n time

t o t h e s a t e l l i t e p a s s . F i e l d s of sa te l l i t e -der ived sur face wind are

compared t o f i e l d s of hourly-observed surface wind in t h ree a r eas .

Finally, rawinsonde and satell i te-derived kinematic parameters are

compared . 6.1 Satel l i te-Derived Winds on Constant-Pressure Surfaces

The bes t s a t e l l i t e -de r ived wind on constant-pressure surfaces is a

geostrophic wind derived from highly smoothed f i e l d s of geopotent ia l

height. A nine-point smoothing routine (Shuman, 1957) w a s a p p l i e d t o t h e

sa te l l i t e -der ived he ight f ie lds over four g r id d i s tances wi th a smoothing

parameter of 0.5. E f fec t s of smoothing sa t e l l i t e -de r ived he igh t f i e lds

before computing geostrophic wind f i e l d s a r e shown in Figs . 19 and 20.

The d i f fe rences between s a t e l l i t e g e o s t r o p h i c wind f i e l d s computed from

smoothed f i e l d s of height and rawinsonde wind f i e l d s a r e shown by s o l i d

l i nes ; s imi l a r d i f f e rences r e su l t i ng from unsmoothed satell i te he ight

f i e l d s a r e shown by dashed lines. Magnitudes of the average and standard

deviat ion of the d i f fe rences between sa te l l i t e -der ived geos t rophic and

rawinsonde wind speeds are decreased when s a t e l l i t e h e i g h t f i e l d s a r e

smoothed i n most areas. Differences i n wind d i r e c t i o n were, in genera l ,

decreased by t h e smoothing process as shown i n Fig. 20.

Average d i f f e rences between geostrophic wind speed computed from

smoothed f i e l d s of sa te l l i t e -der ived he ight and rawinsonde wind speed a re

gene ra l ly pos i t i ve i n t he midd le and upper troposphere as shown i n Fig. 19.

Mean differences range from about -5 t o 5 m s and a r e smallest i n Area I1

(Caribbean) where wind s p e e d s a r e s m a l l a t a l l l e v e l s . Magnitudes of t he

s tandard deviat ion of t he d i f f e rences i n wind speed generally increase

wi th a l t i tude (decrease in p ressure) until t h e l e v e l of the maximum

rawinsonde wind speed is reached. A t t h i s level, magnitudes of the

s tandard deviat ion are approximately 11, 5, 11, and 1 2 m s i n Areas I,

-1

-1

41

100

150

200

100

150

200

250 Q 300

400

500

Y 0)

700

a50

100

150 1 200 -

$ 250 u 300

2 400

b: 500

700

850 -

- s -

-

-5 O 5 10 15 Wind Speed Dif ferencc (m s'l)

a. Area I

-10 -5 0 5 10 15 -1 Wind Speed Difference (m s )

c. Area I11

300

500

700

850

-5 0 5 10 15 Wind Speed Difference (m s )

-1

b. Area I1

-5 0 5 10 15 20 -1 Wind Speed Difference (m s )

d. A r e a I V

"" Unsmoothed - Smoothed

Fig. 19. P r o f i l e s of the average difference and s tandard deviat ion of the differences between satel l i te geostrophic wind speed computed from smoothed and unsmoothed heights and rawinsonde wind speed for Areas I-IV. Differences were computed by subtracting rawinsonde from sa te l l i t e values.

42

100

150

n 200 - 250 5 300 : 400 2 500 700

850

3

-10 0 10 20 30 40 50 60 70 Direction Difference (deg)

a. Area I

100

150 - 2oo : - 250

$ 300 - 2 500 - 3 400 -

700' - 850

-

Avg 0 - -20 0 20 40 60 80 100 120 Direction Difference (deg)

b. Area I1

100

-, 150

r

S 2 O 0 250

a 2 300

- $ 4 0 0

- m

2 500 - 700

- 850

-

L

0 10 20 30 40 50 60 Direction Difference (deg)

c . Area I11

100

150 - 200 250

$ 300 400

500

700

2

u1

85O 1 I 1 -20 -10 0 10 20 30 30 50 60 70

Direction Difference (deg)

d. Area I V

"" Unsmoothed

smoothed

Fig. 20. P ro f i l e s of the average difference and standard deviation of the d i f fe rences between s a t e l l i t e g e o s t r o p h i c wind d i r ec t ion computed from smoothed and unsmoothed heights and rawinsonde wind d i r ec t ion for Areas I - I V . Differences were computed by subtracting rawinsonde from s a t e l l i t e v a l u e s .

43

11, 111, and IV, respectively. The standard deviat ion of the differences

i n wind speed i s between 3 and 1 2 m s i n Areas I, 111, and IV, and

varies from about 2 t o 5 m s i n Area 11.

-1

-1

Average d i f f e rences and standard deviations of the d i f fe rences

between satel l i te geostrophic and rawinsonde wind d i r ec t ions , shown i n

Fig. 20, show t h e l a r g e s t v a r i a t i o n and general ly large magni tudes in

Area I1 where varying wind d i r ec t ions are associated with small wind

speeds. Average differences in direction are between -12 and 41' i n

Area I1 and range from -14 t o 40' i n t h e o t h e r three areas. Magnitudes of

the s tandard deviat ion of t he d i f f e rences i n d i r ec t ion are between about

20 and looo i n Area 11, and range frortl near 15 t o 70' i n t he o the r t h ree

areas. Magnitudes of the standard deviation of the differences i n

d i r e c t i o n are genera l ly smallest near the level of maximum rawinsonde

wind speed.

Sa te l l i t e -der ived and rawinsonde wind f i e l d s are presented in Fig. 2 1

f o r t h e 500-mb l e v e l f o r Area I. The two wind f i e lds have similar flow

pa t te rns wi th cen ters of l a rge d i f f e rences i n wind speed. Both f i e l d s

of wind show ant icyclonic f low and a wind-speed minimum in t he sou theas t e rn

po r t ion of the area, and cyclonic flow and a wind speed m a x i m u m i n t h e

northern portion. The wind speed maximum from sa te l l i t e data (approximately

45 m s ) is loca ted nor theas t of t he m a x i m u m from rawinsonde data (about

35 m s ). Thus, t h e r e are l a rge pos i t i ve d i f f e rences between satell i te-

derived and rawinsonde wind s p e e d s i n t h i s area.

-1

-1

Charac ter i s t ics of t he d i f f e rences between sa te l l i t e -der ived and

rawinsonde wind f i e l d s are as fol lows. Circulat ion pat terns from s a t e l l i t e -

derived geostrophic and rawinsonde wind f i e l d s are similar i n regions of

moderate t o l a r g e wind speeds, but may compare poorly in regions of small

wind speeds. Centers of maximum wind speed i n sa te l l i t e -der ived wind

f i e l d s may be displaced horizontal ly from the cor responding cen ters in

rawinsonde data; a second m a x i m u m i n wind speed may be present in satell i te-

derived winds where none e x i s t s i n rawinsonde data. This also has been

seen by other invest igators (Arnold e t al., 1976) . sa te l l i te-der ived and

rawinsonde winds show good agreement on t h e a l t i t u d e of the j e t s t ream

core, but the je t core from satel l i te data has smaller wind speeds and

less ve r t i ca l shea r of wind speed than are p r e s e n t i n t h e rawinsonde

j e t core.

44

I "

a. Rawinsonde

. . "."I -" I

b. S a t e l l i t e

Fig. 21. Plotted winds and isotach analyses ( m s ) a t 500 m b for the -1

cent ra l Uni ted S ta tes reg ion ( A r e a I ) . Isotachs w e r drawn from exact values and barbs plot ted to the nearest 5 m s . -7

45

U s e of the g rad ien t wind approximation did not improve comparisons

between sa t e l l i t e -de r ived and rawinsonde wind speeds. This is because

the d i f f e rences between satell i te geostrophic and rawinsonde wind speeds

do not correspond t o the curvature of t he s a t e l l i t e -de r ived con tour s .

Areas of large posit ive and negative differences between satell i te-derived

geostrophic and rawinsonde wind speeds are not associated with troughs,

r idges, or any other large-scale pat tern.

6.2. Satel l i te-der ived Surface Wind - Wind speed and direction through the boundary layer t o t h e s u r f a c e

were computed from gridded f i e l d s of geopotent ia l height . The u and v

components of wind were assumed t o vary l inear ly with height above 150 m

t o t h e f i r s t l e v e l of data, and wind speed w a s assumed t o have a

logar i thmic p rof i le below 150 m. Wind d i r ec t ion a t a g r i d p o i n t was

assumed to be constant through the boundary layer. Surface wind speed,

Vs, w a s computed according t o

I n Zs - I n zo vs - ( In zr - In z 1 - vr

- 0

where.Z i s the he ight of t he su r f ace wind, Z is roughness length, and

Z i s a reference height a t which a va lue for wind speed ( V ) i s known. s 0

r r Surface wind speed w a s computed f o r a height of 10 m. A value of

0.5 m was used for roughness length in Areas I and 111, and a value of

0.2 m w a s used in Area 11. These values are in agreement with values

presented by Fied ler and Panof sky (1972) and G a r r a t t (1977) . Fie lds of

satel l i te-der ived geostrophic wind were used t o d e f i n e a reference wind

speed and d i r ec t ion a t each gr id point .

Average d i f fe rences and s tandard deviat ions of t he d i f f e rences

between sa t e l l i t e -de r ived and hourly-observed surface winds are presented

i n Table 6 f o r Area I ( cen t r a l U.S.), Area I1 (Caribbean), and Area I11

(Canada) . The average difference and the s tandard deviat ion of t h e

d i f fe rences between sa t e l l i t e -de r ived and observed surface wind speeds

are smallest in the central Uni ted States where observed wind speeds

were generally between 3 and 8 m s . The large standard deviation of -1

the d i f f e rences i n wind speed i n Canada may be assoc ia ted wi th the l a rge

46

T a b l e 6. Average differences and standard deviations of the d i f fe rences between sa te l l i t e -der ived (S) and hourly-observed (0) surface winds (S-0) for three regions.

Region Speed (m s ) -1 Direction (deg)

Avg. Std. Dev. Avg. Std. Dev. - Central United States -0.3 2.1 16 34

Caribbean 1.5 2.8 2 1 66

Canada 0.9 4.3 30 28

wind speeds and the intense low-pressure center in the area. The

di f fe rences between satell i te-derived and observed wind speeds in the

Caribbean are larger thaq expected in this region of very low wi?d speeds.

The magnitude of the s tandard deviat ion of the differences in wind

d i r ec t ion is l a r g e s t i n t h e Caribbean where surface winds were l i g h t and

varidble. The magnitcde of the standard deviation of the d i f fe rences in

surface wind d i r e c t i o n is smal les t in Canada where the s a t e l l i t e -de r ived

f low pat tern i s similar to the well-organized observed f low pattern.

Fields of sa te l l i t e -der ived and observed surface winds are presented

i n Fig. 22 f o r Area I. Both f i e l d s of wind ind ica te an t icyc lonic f low in

the southeastern port ion of the region, weak cyclonic f low in the northern

portion, and strong cyclonic flow around the surface low-pressure center in

Oklahoma. Magnitudes of the d i f fe rences in wind speed are less than 3 m s

a t most gr id po in ts . Observed surface winds acce le ra t e as they c ross the

isobars toward lower pressure. This acceleration w a s not taken into

account i n the computation of sa te l l i t e -der ived sur face wind speeds and

-1

l eads t o nega t ive d i f f e rences i n wind speed (satel l i te values too small)

near Oklahoma and the Great Lakes.

6.3 Comparisons of Satel l i te and Rawinsonde-derived Kinematic Parameters

Kinematic parameters were computed from gr idded f ie lds of rawinsonde "- I

and s a t e l l i t e d a t a f o r Areas I-IV. Horizontal advection of temperature,

t h e v e r t i c a l component of r e l a t i v e v o r t i c i t y , and the horizontal advect ion

of abso lu te vor t ic i ty were computed. The rawinsonde calculations used

f i e l d s of temperature and wind from rawinsonde measurements, while t h e

47

I ". a. Observed

s a t e l l i t e c a l c u l a t i o n s used f i e l d s of temperature and geostrophic wind

from s a t e l l i t e d a t a .

Rawinsonde and sa t e l l i t e -de r ived f i e lds of temperature advection

a re s imi l a r a t 850 and 500 mb. A s shown i n Fig. 23, rawinsonde and

sa t e l l i t e -de r ived f i e lds of temperature advection a t 850 mb ind ica te

cold-air advection over northern Wisconsin and warm-air advection over

northeastern Oklahoma. Magnitudes of warm-air advection a r e n e a r l y t h e

same for bo th types of data, while satell i te-derived magnitudes of cold-

air advection over Wisconsin are smaller than the rawinsonde values.

F ie lds of horizontal advection of temperature for Areas 11-IV (not

shown) i n d i c a t e t h a t s a t e l l i t e d a t a a r e c a p a b l e of depict ing centers of

pos i t i ve and negative temperature advection for each of the synoptic

conditions considered i n th i s s tudy .

There is l i t t l e correspondence between the rawinsonde and s a t e l l i t e -

der ived f ie lds of r e l a t i v e v o r t i c i t y a t 500 mb. Centers of r e l a t i v e

v o r t i c i t y from t h e two da ta s e t s a r e gene ra l ly of opposite sign i n

Areas I, 11, and IV. Fie lds of r e l a t i v e v o r t i c i t y computed from the

two types of data are s imilar only i n Canada where t h e 500-mb flow was

strong and cyclonic. Fields of sa te l l i t e -der ived advec t ion of absolute

v o r t i c i t y a t 500 mb are diss imilar to corresponding rawinsonde f ie lds in

each of the four a reas .

49

a. Rawinsonde

b. S a t e l l i t e

Fig. 23. Fie lds of horizontal advect ion of temperature (10 OC S-l) -6

at 850 mb for the central Uni ted States region (Area I).

50

7 . COMPARISONS BETiEEN SIMULTANEOUS TIROS-N AND RAWINSONDE DATA FOR

2100 GNT ON 10 APRIL 1979

The ana lys i s of atmospheric structure determined from quan t i t a t ive

satel l i te data has been extended t o i n c l u d e a case with simultaneous

rawinsonde and TIROS-N sounding data. This research has two object ives .

The f i r s t c b j e c t i v e is t o determine the l imitat ions of TIROS-N sounding

da ta for the purpose of determining the atmospheric struckure in a

meteorological ly act ive area. The second objective of t h i s r e s e a r c h i s

t o a id i n t he eva lua t ion of t h e r e s u l t s o b t a i n e d f o r Areas I - V I 1 with

Nimbus-6 and time-interpol-ated rawinsonde data. Simultaneous TIROS-N

and rawinsonde soundings provide an opportunity t o compare satel l i te and

rawinsonde data without the r isk of i ncu r r ing e r ro r s from a time-

interpolat ion process .

Except fo r t he l ack of a t ime-interpolation process, a l l procedures

used in the ana lys i s of d a t a i n t h e AVE-SESAME area a re i den t i ca l . t o

those followed for Areas I - V I I . Satell i te data, rawinsonde data, and

synopt ic condi t ions for the AVE-SESAVfl area were described in Section 2.

7 . 1 Analysis of Discrepancies Between Rawinsonde and TIROS-N P r o f i l e

Parameters

For the purpose of comparisoa, TIROS-N soundings were paired with the

closest rawinsonde soundings. N o t a l l sa te l l i t e data were used siilce

t h e r e were more satell i te than rawinsonde soundings. The 20 pa i r ings of

s a t e l l i t e sounding locations and rawinsonde s t a t i o n s a r e snob-n in Fig. 24.

Seven parameters were considered in this s tudy: temperature ,

dew-point t enpe ra twe , mixing r a t io , t h i ckness , l apse r a t e of texperature ,

p rec ip i t ab le w a t e r , and s tabi l i ty . Discrepancies between s a t e l l i t e and

rawinsonde data for a l l seven parameters w e r e computed by subtract ing

rawinsonde from sa te l l i t e values and were analyzed in the sane manner as

those obtained for Areas I - V I I .

7 .1 .1 Temperature

T a b l e 7 shows t h e mean and standard deviation of t h e mearr discrepancy,

the absolu te m e a n discrepancy, and the root-mean-square discrepancy (RVSD)

between TIROS-N and rawinsonde temperatures. The statistics were obtained

from t h e lumped discrepancies for a l l leve ls repor ted from each s ta t ion in

51

Fig. 24. Pa i r ings of s a t e l l i t e sounding locations and rawinsonde s t a t i o n s a t 2100 GMT on 10 April 1979.

The region t o p r o v i d e a s ing le set of c r i t e r i a by which t o judge the

r e s u l t s of the comparisons. The mean discrepancy of -0.5OC i s of

opposite sign of those found in previous areas (see Table 2 ) . This

ind ica tes TIROS-N-derived temperatures contain a n e g a t i v e b i a s r e l a t i v e t o

rawinsonde-derived temperatures. The mean RMSD of 1.8OC is smaller than

t h a t found i n most of the previous areas (Table 2 ) .

The m e a n s and standard deviations of temperature discrepancies for

Table 7. M e a n (x) and standard deviation (a), lumped f o r a l l l e v e l s repor ted for each s ta t ion and f o r a l l s t a t i o n s i n t h e SESAME region, of t h e - mean discrepancy (8) , t he abso lu t e mean discrepancy ( 16 1 ) , and the root-mean-square discrepancy (RMSD) , in degrees Cels ius , between TIROS-N and rawinsonde temperatures and dew po in t s [6 E (T - TR) ] .

S

Temperature Dew-Point Temperature Station Pairs

6 181 RMSD 6 181 RMSD - -

- X -0.5 1.5 1.8 -2.7 9.4 10.9

0 0.5 0.4 0.5 8.4 4.2 4.3 20

~

52

a l l t h ree l aye r s are shown i n T a b l e 8, and the cumulative frequency

d i s t r i b u t i o n s p l o t t e d on probabili ty paper are shown i n Fig. 25.

Table 8. Means and standard deviations of discrepancies (S-R) between TIROS-N and rawinsonde data for selected parameters by layer f o r t h e SESAME region a t 2100 GMT on 10 A p r i l 1979.

1000-500 mb 500-300 mb 300-100 mb

Mean -0.8 0.4 -0.6 Temperature Standard deviation 1.8 1.5 1.8

(OC) No. of da ta po in ts 63 60 96

Mean -3.3 -2.3 Dew point Standard deviation 9.1 12.9

(OC) No. of da t a po in t s 63 60

Mean -0.3 -0.1 0.3 Lapse r a t e Standard deviation 0.5 0.7 0.8

( " C / h ) No. of data poin ts 43 59 76

Mean -0.9 0.0 Mixing r a t i o Standard deviation 2.0 0.5

(gm/kgm) No. of da ta po in ts 63 60

Mean Precipitable Standard deviation water (mm) No. pa i r ed p ro f i l e s

-1.2 2.3 20

. . . . . - . .

The mean d i sc repanc ie s l i s t ed i n Table 8 i n d i c a t e t h a t t h e r e i s a

negative bias between the satel l i te and rawinsonde temperature data in

t h e 1000 t o 500- and 300 t o 100-mb layers. The nega t ive b i a s i n s a t e l l i t e -

derived temperatures f o r t h e 300 t o 100-mb layer w a s not present in any

of t h e areas previously studied (see Table 3) and i s pa r t i a l ly r e spons ib l e

for the nega t ive mean discrepancy for t h e lumped data (Table 7). Magnitudes

of the standard deviation of temperature discrepancies range from 1.5OC i n

the middle l ayer to 1.8OC i n t h e upper and lower l aye r s and are s i m i l a r t o

those found i n t h e f i r s t c e n t r a l United States case (Area I ) . Except f o r

53

I 0. I I I I 1

0. I I 10 30 5 0 70 90 Cunulative Probabili ty ( % )

99 9 9.9

Fig. 25. Cumulative probabili ty frequency distributions of tezqeracu-re discrepancies within the layers 1000 t o 500 mb, 500 t o 300 mb, and 300 t o 100 m b f o r t h e AVE-SESAPIE area.

t h e second Canada and Caribbean areas, the standard deviation of temperature

discrepancies is smallest in the middle l ayer for a l l areas studied.

The cumulative frequency distributions shown in F ig . 25 are approxi-

mately normal (straight l ines) except near the extremes. The small sample

s i z e i s inadequate for defining the exLtcemes of the d i s t r ibu t ions . - The

tendency for the cumulative f r equency d i s t r ibu t ions t o be s t r a igh t l i nes

when p lo t ted on probabi l i ty paper suggests that the discrepancies between

TIROS-N and rawinsonde temperatures are due t o random e r ro r s .


S t a t i s t i c s €or the ensemble of discrepancies between satel l i te and

rawinsonde dew-point temperatures also a r e shown i n Table 7. TIROS-N

soundings of dew-point temperature a r e n o t as r e l i a b l e as those of

temperature. The mean RMS discrepancy for the SESAME area, 10.9OC, i s

larger than those found i n a l l but one of t h e areas previously studied.

54

I Discrepancies in dew-point temperature were examined f o r t h e 1000

t o 500- and 500 t o 300-mb layers . Means and s tandard deviat ions of t h e

discrepancies within the t w o l aye r s are shown i n Table 8, and cumulative

frequency dis t r ibut ions are shown i n Fig. 26. The mean d i f fe rence i s

smaller in t he h ighe r l aye r t han i n t he lower layer, a r e su l t oppos i t e

of those found i n a l l areas studied previously, except for the second

Canadian area. The standard deviation of t he d i sc repanc ie s i n dew-point

temperature are smal le r in the l o w e r layer than in the upper layer . This

ag rees w i th t he r e su l t s found i n most of the previous areas. The p lo t t ed

cumulative frequency distributions for dew-point temperature discrepancies

(Fig. 26) are not as s t r a i g h t as those for temperature discrepancies (Fig.

25), but may be considered as s t r a i g h t l i n e s as a f i r s t approximation.

Fig. 26. Cumulative probabili ty frequency distributions of dew-point temperature discrepancies within the layers 1000 t o 500 mb and 500 t o 300 m b for t h e AVE-SESAME area.

55

7 -1.3 Thickness

Layer th ickness d i screpancies , normal ized to un i t s of m km , were -1

s t r a t i f i e d i n t o three layers . Means and s tandard deviat ions of thickness

discrepancies are p resen ted i n T a b l e 9. Values of the s tandard deviat ion

of discrepancies range from 4.7 t o 6.3 m km and increase wi th a l t i tude .

Cumulative probabi l i ty curves for normalized discrepancies in thickness

(not shown) are approximately s t ra ight l ines .

-1

T a b l e 9. Means and standard deviations of normalized discrepancies i n t h i c k n e s s f o r t h e l a y e r s 1000 t o 500 m b ( A ) , 500 t o 300 -1 m b ( B ) , and 300 t o 100 mb ( C ) f o r t h e AVE-SESAME a rea (m km ) .

A B C

Mean -3.1 0.3 -1.9

S t . Dev. 4.7 5.4 6.3

No. of Data 43 60 96

7.1.4 Mixing Rat io

Mixing r a t i o v a l u e s were obtained from dew-point temperature data

p lo t t ed on skew T-log p diagrams for rawinsonde and TIROS-N soundings.

The r e s u l t s of comparisons between s a t e l l i t e and rawinsonde-derived

mixing r a t i o s are presented in Table 8. The mean and standard deviation

of the d i screpancies in the lower l ayer are grea te r than those found i n

the upper layer for the AVE-SESAME a rea . These r e su l t s a r e i n agreement

with those found for a l l previous areas studied (Table 3 ) . Magnitudes of

the s tandard deviat ion of the d i screpancies in mix ing ra t io a re 2.0 and

0.5 g kg for the lower and upper layers, respectively. These values

a r e similar t o t h o s e found i n t h e o t h e r areas.

-1

7.1.5 Precipitable Water

The mean and s tandard deviat ion of the discrepancies in precipi table

water =e presented in Table 8. The r e s u l t s i n d i c a t e t h a t TIROS-N

soundings yield values of precipi table water which a r e smaller than those

from rawinsonde data. The s tandard deviat ion of 2.3 mm is of approximately

the same magnitude as those found in the other areas (see Table 3).

56

7.1.6 Lapse Rate of Temperature

Lapse rates computed from sa te l l i t e and rawinsonde data were normalized

t o u n i t s of OC Ian-'. Discrepancies in lapse rate were s t r a t i f i e d i n t o

th ree l aye r s : 1000 t o 500 mb, 500 t o 300 mb, and 300 to 100 mb. Resul ts shown i n Table 8 ind ica te tha t sa te l l i t e -der ived lapse rates

have a negat ive b ias in the lower t w o l aye r s and a p o s i t i v e b i a s i n t h e

upper layer. Magnitudes of the standard deviation of the d i screpancies

range from 0.5 t o 0.8OC lan and are general ly smaller than those found

in p rev ious areas ( T a b l e 2) . When comparing normalized r e s u l t s from t h e

SESAME area with those from previous areas, the depth of the layer through

which the lapse rate i s computed must be considered. Because soundings

f o r t h e f i r s t f o u r areas contained 2 1 l e v e l s of da t a and f o r t h e AVE-SESAME

area on ly 10 l eve l s , r e su l t s can no t be s t r i c t ly compared. Resul ts from

Areas V - V I 1 may be compared with those from the present s tudy s ince

soundings fo r t hese a r eas a l so con ta ined 10 l eve l s of da ta . Magnhtudes

of the s tandard deviat ion of d i s c r e p a n c i e s i n l a p s e r a t e are smaller and

have a smaller r ange i n t he AVE-SESAME area t h a n i n Areas V - V I I .

-1

7.1.7 S t a b i l i t y

Showalter and Vert ical Totals indexes computed from TIROS-N and

rawinsonde data and the discrepancy for each station pair are shown i n

Tables 10 and 11, respec t ive ly . The mean and standard deviation of the

d iscrepancies a l so are presented for each index.

A l l Showalter Indexes computed from s a t e l l i t e d a t a were p o s i t i v e ; t h i s

a l s o w a s t rue in each of the other areas s tudied. Smal le r percentage errors

i n t h e mean and s tandard deviat ion of discrepancies were found f o r t h e

v e r t i c a l to ta ls index than for the Showalter index. The mean and standard

deviat ion of discrepancies are 4 . 1 and 4.6 for the Showalter index, and

-0.9 and 2.2 f o r t h e v e r t i c a l t o t a l s i n d e x . These r e s u l t s are s i m i l a r t o

those found for Areas I-IV.

7 .2 Analysis of Discrepancies Between Rawinsonde - and TIROS-N Data on

constant-Pressure Surfaces

7.2.1 Temperature

P ro f i l e s of the average and s tandard deviat ion of dif ferences in

temperature are shown i n F i g . 27. The average difference increases from

57

Table 1 0 . D i s c r e p a n c i e s i n t h e S h o w a l t e r Index d e r i v e d f r a n TIROS-N a n d r a w i n s o n d e d a t a for the AVE-SESAME area.

". ~

S t a t i o n No. S a t e l l i t e Bawinsonde Discrepancy

353 553 53 2 433 260 456 255 363 232 229 261 4 51 23 5 240 340 247 265 349 3 27 562

6.6 8.9

13.8 13.1 1.8 7.3 0.5 4.7 4.4 7.8 7.6 2.0 5.8 7.0 6.6 0.3 6.8 9.0 7.5 9.2

-2.9 7.3

15 .5 15.3 -4.3

3.0 -4.5

3.6 4.4

-0.8 -3.9

2.1 -2.3 -0.2 -2.0 -2.6

1.0 -1.2 11.7 . 10.3

9.5 1.6

-1.7 -2.2 6.1 4.3 5.0 1.1 0.0 8.6

11.5 -0.1 8.1 7.2 8.6 2.9 5.8

10.2 -4.2 -1.1

Tab le 11. D i s c r e p a n c i e s i n t h e V e r t i c a l Totals I n d e x d e r i v e d from TIROS-N and r a w i n s o n d e d a t a for t h e AVE-SESAME area.

~ ~~

S t a t i o n No. S a t e l l i t e -&winsonde Di sc repancy

3 53 553 532 43 3 260 456 255 363 232 229 261 451 235 240 340 247 265 349 327 562

24.7 23.8 21.0 22.4 27.3 24.8 25.7 28.6 22.2 25.2 30.4 27.7 23.7 24.4 24.0 28 .O 31.7 24.4 26.6 23.5

27.3 23.3 18.0 23.5 27.9 24.9 26.9 28.4 24.4 25.3 31.4 30.0 26.6 28.8 26.3 27.0 36.4 26.9 24.9 19.5

-2.6 0.5 3.0

-1.1 -0.6 -0.1 -1.2

0.2 -2.2 -0.1 -1.0 -2.3 -2.9 -4.4 -2.3

1 . 0 -4.7 -2.5

1 .7 4.0

''-22 -0.9 S t a n d a r d D e v i a t i o n 2.2

58

Temperature Difference ("C)

Fig. 27. P r o f i l e s of average and standard deviation of differences between satel l i te and rawinsonde temperatures (satel l i te minus rawinsonde) for the AVE-SESAME area.

approximately -1.5OC a t 850 mb t o n e a r l y 0.7OC a t 300 m b , then decreases

t o -1.3OC a t 100 mb. The m a x i m u m standard deviation of the d i f fe rences

is about 1.9OC and occurs a t 850 and 200 mb. The s tandard deviat ion i s

about 1 . 2 O C a t 500, 300, and 100 mb. Results from previous areas a l s o

ind ica ted tha t re la t ive ly l a rge magni tudes of the s tandard deviat ion

occur near the tropopause and the ground. Comparison of r e s u l t s shown

i n Table 1 2 with those for previous areas (Table 5) indicates that the

magnitude of the s tandard deviat ions of differences between TIROS-N and

simultaneous rawinsonde temperatures are similar t o t h o s e between Nimbus-6

and time-interpolated rawinsonde temperatures.


P r o f i l e s of the average and s tandard deviat ion of t he d i f f e rences

between TIROS-N and rawinsonde dew-point temperatures are shown i n F i g . 28.

The average difference is negative a t a l l l e v e l s which i n d i c a t e s t h a t

satell i te dew-points are, on the average, lower than rawinsonde values.

The s tandard deviat ion of t he d i f f e rences r anges from about 5 t o 10°C and

is considerably larger than the corresponding values for temperature

presented i n Fig. 27. Comparison with past r e s u l t s f o r dew-point

temperature ind ica t e s that the average difference general ly i s of opposite

s ign t o those ob ta ined in p rev ious areas and the s tandard deviat ion is

l a r g e r f o r t h e AVE-SESAME area.

59

Table 12. Means and standard deviations of discrepancies between gridded satell i te and rawinsonde parameters on selected constant pressure s u r f a c e s f o r t h e AVE-SESAME area a t 2100 GMT on 10 A p r i l 1979.

700 m b 500 mb 300 m b

Temperature ("C) Mean -1.4 -0.6 0.7 Standard Deviation 1.6 1.3 1 . 2 Approx. Magnitude -3/10 -18 -4O/-48

Dew-point Mean -3.8 -5.4 -6.3

Temperature Standard Deviation 10.2 5.4 9.9 (OC) Approx. Magnitude -2/-20 -2O/-36 -52

~~

Lapse ,Pate of Mean -0.2 -0.3 0.1 Temperature Standard Deviation 0.5 0.3 0.8 (OC/km) Approx. Magnitude 4.5/8.0 7.5 7.2

Magnitude of Horizontal Mean 0.3 2 .1 0.0 Gradient of Standard Deviation 4.3 3.2 5.1 Temperature Approx. Magnitude 1/25 1/15 1/13 (oc/ looo km)

~.

Geopotential Mean -19.2 -32.3 -27.8 Height (m) Standard Deviation 18.2 27.0 42.6

Approx. Magnitude 3000 5600 9300

Geo. u-comp. Mean -0.7 0.9 3.9 wind ( m / s ) Standard Deviation 5.4 7.0 10.9

~ ~ ""

Approx. Magnitude -6/30 -2/40 -7/50

Geo. v-comp. Mean 2.5 4.0 6.0 Wind ( m / s ) Standard Deviation 4.5 7.0 11.1

Approx. Magnitude 4/26 5/36 10/50

Scalar Wind Mean 2.9 4.6 8.1 Speed ( m / s ) Standard Deviation 4.1 7.0 11.5

Approx. Magnitude 10/30 15/50 15/7 0

Wind Mean -8.7 -4.8 -1.7 Direction (deg) Standard Deviation 20.2 17.6 17.0

Approx. Magnitude 2 00 210 23 0

N o . of da t a po in t s 95

60

100-

150 - 200 -

$ 250- a, 300-

400 - a,

PI &I 500 - 700 - 850 -

-10 -5 0 5 10 Dew-point Temperatur e Di f fe r ence ( "C)

Fig. 28. P ro f i l e s of average and s tandard deviat ion of d i f fe rences between s a t e l l i t e and rawinsonde dew-point temperatures ( s a t e l l i t e minus rawinsonde) f o r t h e AVE-SESAME area.

7.2.3 Lapse R a t e and Horizontal Gradient of Temperature - Prof i l e s of the average and standard deviation of d i f f e r e n c e s i n

lapse rates a r e shown in F ig . 29. The average difference curve shows t h a t

lapse rates from TIROS-N da ta are smaller than those from rawinsonde da ta

a t l e v e l s below 300 mb and are larger above 300 mb. Average d i f fe rences

vary from about -0.4 t o 0.7OC km-l which i s a larger range than w a s found

in p rev ious areas. The standard deviation ranges from nea r ly 0 .3 t o 0.9OC km

with a maximum magnitude a t 250 mb.

-1

Ver t i ca l d i f f e rence p ro f i l e s fo r t he ho r i zon ta l g rad ien t of temperature

a r e shown i n Fig. 30. The average d i f fe rence var ies from about -2 t o 2OC

(1000 km)-' with gradients from satel l i te data being larger on constant-

pressure surfaces between 700 and 300 mb and above 200 mb. Magnitudes of

the s tandard deviat ion range from approximately 3.2 t o 6.5OC (1000 km) and

are similar t o t h o s e found i n t h e f i r s t c e n t r a l U n i t e d S t a t e s a r e a .

-1

7.2.4 Geopotential Height

Ver t ica l d i f fe rence p rof i les for geopotent ia l he ight a re shown i n

Fig. 31. Average differences decrease from about -7 m a t 850 mb t o n e a r l y

-38 m a t 100 mb, indica t ing that sa te l l i t e -der ived geopotent ia l he ights

61

100

- - Avg "' 0 k 500 700

- $ 400

- a, 300

- 9 250

- 200

- 150

-

h

v

L4

a,

850 L -0.5 0 0 .s 1.0

Lapse Rate ("C km-')

Fig. 29. Prof i les of average and standard deviation of differences between sa t e l l i t e and rawinsonde ve r t i ca l l apse rates f o r t h e Am-SESAME area.

100

150 - 200

2 250

- h - v

$ 300 -

400 - a, k

500

700

850 -

- - Av9

- 2 0 2 4 6 8 Horizontal Temperature Gradient ( "C/1000 km)

Fig. 30. Profiles of average and standard deviation of differences between satel l i te and rawinsonde horizontal temperature grad ien ts for the AVE- SESAME area.

-40 -20 0 20 40 60 Geopotential Height (m)

Fig. 31. Prof i les of average and standard deviation of differences between sa t e l l i t e and rawinsonde geopotentia.1 heights for t he Am-SESAME area.

are smller, on the averaga, than those from rawinsonde data a t a l l

levels . This is due t o the nega t ive b i a s i n TIROS-N temperatures and

dew-point temperatures relative t o rawinsonde data. Magnitudes of t h e

s tandard deviat ion increase from nezrly 16 m a t 850 m b t o about 43 m a t

300 mb, then decrease to approximately 29 m a t 100 mb. The range

in the s tandard devia t ion is similar t o t h o s e found i n o t h e r areas.

7.2.5 Geostrophic Wind

P ro f i l e s of t he d i f f e rences between geostrophic winds computed from

rawinsonde and TIROS-N geopotent ia l heights are presented in F ig . 32 f o r

t h e u and v component wind speeds, scalar wind speed, and wind d i rec t ion .

Average d i f fe rences between t h e component wind speeds zre less than 6 m s

a t a l l a l t . i tudes and are generally positive. Magnitudes of the s tandard

devia t ion of t he d i f f e rences i n component wind speeds range from about 4

t o 1 4 m s and are la rges t near the l eve l o f the tropopause. Average d i f -

ferences between geostrophic scalar wind speeds range from about 3 t o 8 m s , indicat ing that geostrophic wind speeds compute6 from sa t e l l i t e -de r ived

h e i g h t f i e l d s are Larger on the average, than those from rawinsonde data

a t a l l levels. Standard deviations increase fron about 3.5 111 s a t 850 mb

t o 13.0 m s a t 250 mb, then decrease to nearly 10.5 m s a t 100 mb. -1 -1

Magnitudes of the average difference i n wind d i r e c t i o n are less than 10' a t

a l l levels , whi le the s tandard deviat ion of the d i f fe rences i n d i r e c t i o n

ranges frorn about 16 t o 28'.

-1

-1

-1

-1

Comparisor, of these rcsu l t s wi th those from previous areas shows that

d i f f e r e n c e s i n component and sca la r wind speeds i n t h e AVE-SESAME area a r e

similar t o those i n o the r a r eas . However, the magnitude and range of t h e

average and standard deviation of the d i f fe rences in geos t rophic wind

d i r e c t i o n a r e s i g n i f i c a n t l y smaller i n t h e AVE-SESAME area. Mean d i f fe rences

and standard deviations of t h e d i f f e x e n c e s i n wind direction ranged from

about -30 t o 0' and 30 t o 80°, r e s p e c t i v e l y , i n t h e f i r s t c e n t r a l U n i t e d

States area.' The improved r e su l t s fo r geos t roph ic wind d i r e c t i o n i n t h e

present area probably are due t o the, synopt ic condi t ions in the area o r t h e

use of simultaneous rawinsonde and satel l i te data.

1

P r o f i l e s of the differences between observed wind and satel l i te-der ived

geostrophic wiEd are presented i n Fig. 33. AvesaSe differences between the

component wind speeds range from about -2 to 7 in s , while the standard

deviations vary from about 4 t o 14 m s . Average d i f f e s e n c e s i n scalar

-1

-1

63

100

150 - 200

3 250

- - -

v

400 a, &4

500 - 700 - 850 -

L

100 - 150 - 200

# 250

$300

$400

- h

" e - 2

E 500 - 700 - 850

Direction (des)

Fig. 32. Prof i les of average and standard deviation of differences between geostrophic winds computed from rawinsonde and sa te l l i t e geopotent ia l he ights for the AVE- SESAME area. Differences were computed by subtracting rawinsonde from satell i te values.

100

150 - 200 - 1 9 250 a, 300 Ll 1 - ar

- v -

400

500

700

- -

850 - Avg a

-5 0 5 10 15 -1

u ( m s 1

-10 0 10 20 30 Direction (deg)

Fig. 33. Prof i les of average and standard deviation of differences between rawinsonde winds and satell i te-derived geostrophic winds fo r t he AVE-SESAME area. Differences w e r e computed by subtracting rawinsonde from sa te l l i t e values.

wind speed a r e p o s i t i v e a t most l e v e l s which ind ica t e s that s a t e l l i t e -

derived geostrophic wind speeds are larger , on the average, than rawinsonde

wind s2eeds. The s tandard deviat ion of t he d i f f e rences i n s ca l a r wind

speed increases from about 4 m s a t 850 m b t o approximately 1 2 m s -1 -1

between 300 and 200 mb, then decreases to near 7 . 5 m s a t 100 mb. Average

d i f f s r ences i n wind d i rec t ion range from about -10 t o lo', while the

s tandard deviat ion is re l a t ive ly cons t an t a t about 18'. Results obtained

fo r t he p re sen t a r ea fo r wind speed a re s imi l a r t o t h o s e i n t h e f i r s t

United States area. Differences i n wind d i r e c t i o n a r e s i g n i f i c a n t l y

smaller i n t he AVE-SESAME area .

-1

66

8 . SUMMARY A N D CONCLUSIONS

8.1 Summary

The c a p a b i l i t i e s of Nimbus-6 and TIROS-N satel l i te sounding data for

use in determining atmospheric structure have been investigated for

several geographic areas. A n evaluation of t h e a b i l i t y of t h e sstell i te

d a t a t o d e p i c t s t r u c t u r a l f e a t u r e s of the atmosphere w a s base6 on

comparisons between sa te l l i t e and rawinsonde data. Nimbus-6 d a t a were

compared t o time-interpolated rawinsonde data, aEd simultaneous TIROS-N

and rawinsonde data were campared. Two approaches t o t h e a n a l y s i s and

comparison of sa te l l i te and rawinsonde data w e r e followed: 1) di f fe rences

between paired soundings of sa te l l i te and rawinsonde data were computed,

and 2) da t a f rom the s a t e l l i t e and rawinsonde soundings for selected

constant-pressure surfaces were gridded and values from the two sets of

d a t a were compared a t the g r id po in ts .

8.2 Conclusions

The following conclusions were reach4 from t h e r e s u l t s of t h i s

research:

(1) The approximate mean RMS of t h e d i s c r e p a n c i e s f o r p r o f i l e pairs

between Nimbus-6 and time-interpolated rawinsonde data f o r seven parameters

and a l l seven areas are the fol lowing:

(a) Temperature: 2OC

(b) Dew-point temperature: 7.5OC

(c) Layer thickness: 7 m km-l

(d) Mixing r a t i o : 1.34 g kg -1

(e ) Prec ip i tab le water: 0.23 c m

( f ) Lapse r a t e of temperature: 1.10~ km-l

(9) A l l Showalter indexes derived from s a t e l l i t e d a t a a r e p o s i t i v e ,

and t h e v e r t i c a l to ta ls index i s within 5% of and smaller than those computed

from rawinsonde data.

( 2 ) Cumulative frequency distributions show tha t d i sc repanc ie s between

Nimbus-6 and rawinsonde data can be represented by a normal d i s t r ibu t ion .

(3 ) For temperature and temperature-related var iables , there i s a

strong correspondence between gridded fields of rawinsonde and Nimbus-6

data . Temperature d i f fe rences are s igni f icant on ly in reg ions of s t rong

v e r t i c a l or horizontal gradients . In cross sec t ions and constant-pressure

67

cha r t s , t he satell i te da ta y i e ld s imi l a r pa t t e rns t o rawinsonde data,

e x c e p t t h a t f r o n t a l c o n t r a s t s are somewhat smoothed so t h a t g r a d i e n t s

behind f ronts are not quite as s t r o n g i n t h e sa te l l i t e data. Differences

between sa te l l i t e and rawinsonde temperatures t e n d t o b e l a r g e s t n e a r t h e

tropopause and the ground. Lapse rate of temperature, along with

temperature, i s use fu l fo r de t e rmin ing f ron ta l l oca t ions from s a t e l l i t e

data .

(4 ) For gridded f ields of dew-point temperature and other measure-

ments of moisture, the Nimbus-6 soundings present a smoothed vers ion of

rawinsonde soundings. Examination of dew-point temperature i t s e l f seems

t o y i e l d poor r e s u l t s i n terms of t he dep ic t ion of f r o n t a l c o n t r a s t s and

i n t e r m s of quan t i t a t ive d i f f e rences between satel l i te and rawinsonde

values. Equivalent potential temperature, which combines temperature and

moisture measurements, i s shown t o b e a be t t e r va r i ab le fo r dep ic t ing

f ron ta l l oca t ions .

(5) Differences between rawinsonde and sa t e l l i t e -de r ived f i e lds of

geopotent ia l height tend t o increase toward the tropopause and decrease

s l i g h t l y above t h a t level.

(6) Resu l t s i nd ica t e t ha t t he bes t s a t e l l i t e -de r ived wind on constant-

p ressure char t s i s a geostrophic wind derived from highly smoothed f i e l d s

of geopotential height. Satell i te-derived winds computed i n t h i s manner

and rawinsonde winds show similar c i r c u l a t i o n p a t t e r n s e x c e p t i n areas of

small height gradients. Magnitudes of the s tandard deviat ion of t h e

differences between satel l i te-der ived and rawinsonde wind speeds range

from about 3 t o 1 2 m s on constant-pressure charts and peak a t t h e

je t -s t ream level .

-1

(7) F ie lds of sa te l l i t e -der ived sur face wind computed with the

logarithmic wind l a w agree w e l l w i th f i e lds of observed surface wind i n

most regions. Magnitudes of the standard deviation of the differences in

sur face wind speed range from about 2 t o 4 m s , and sa te l l i t e -der ived

surface winds are able to dep ic t f l ow ac ross a co ld f ront and around a

low-pressure center.

-1

(8) Results obtained from the comparison of simultaneous TIROS-N and

rawinsonde data are similar t o t h o s e found f o r Nimbus-6 and t ime-interpolated

data . The on ly s ign i f i can t change i n t he r e su l t s w a s t h a t found f o r t h e

differences between satellite-derived and rawinsonde wind d i r ec t ion .

68

Magnitudes of the average and standard deviation of the d i f fe rences

between TIROS-N and rawinsonde wind d i r e c t i o n s are approximately half

as l a rge as the corresponding differences for Nimbus-6 and rawinsonde

da ta . The improved r e s u l t s f o r wind d i rec t ion wi th TIROS-N da t a may

be due to t he synop t i c cond i t ions i n t he a r ea o r t he u se of simultaneous

rawinsonde and satell i te data .

Arnold, J. E., J. X. Scoggins, and H . E. Fuelberg, 1976: A comparison between Nimbus-5 THIR and ITPR temperatures and derived winds with rawinsonde data ob ta ined i n t he AVE I1 experiment. NASA Contractor Report CR-2757, 76 pp.

Barnes, S. L. , 1964: A technique for maximizing detai l in numerical weather map ana lys i s . - J. 9 p l . - Meteor. , 3 , 396-409.

F ied ler , F., and H. 4. Panofsky, 1972: The geos t rophic d rag coef f ic ien t and the effect ive roughness length, Ouart . J. Roy. Meteor. SOC., 98, 213-220.

”- -

Garratt, J. R . , 1977: Review of cirag coefficients over oceans and continents. ”- - Mon. Wea. Rev., 105, 915-929.

Grody, PJ. C. , C. M. Hayden, ail6 W. C . Shen, 1979: Typhoon June Winds Estimated From Scanning Microwave Spectrometer Measurenents a t 55.45 GHz. .T. ceophys. R e s . , 8 4 , 3689-3695. - ”

Hanel, R . , and B. J . Conrath, 1969: Interferometer experiment on Nimbus 3: Preliminary Results. Science, 165, i258-1260. -

Hil lger , D. W . , and T. H. Von der aaar, 1977: Deriving mesoscale temperature and x o i s t i l r e f i e l d s from s a t e l l i t e radiance measurements over t h e United States. J. Appl. .h4eteor., 16, 715-726.

” - Horn, L. H. , R. A. Petersen, and T. M. Whittaker, 1976: Intercoinparisons of

data derived from Nimbus 5 t e m p r a t u r e p r o f i l e s , rawinsonde observations and i n i t i a l i z e d LFM model f i e l d s . Non. Wea. Rev., 104 , 1363-1371.

”- -

Kapela, A . F . , and L. E. Horn, 1975: Ninbus-5 s a t e l l i t e soundinqs i n a s t rongly barocl inic region. Pleteorological Applicat ions of Satel l i te Indirect Soundings, Project Report, PjOAA Grant 04-4-158-2, Department of Keteorology, University of Wisconsin, ls?adison, 1-19.

Petersen, R. A . , and L. B. Horn, 1977: An evaiuation of 500-mb height and geostrophic wind f i e l d s d e r i v e d from Nimbus-6 soundings. Bull. in. Meteorol. SOC., 58, 1195-1201.

”

”

Shuman, F. G . , 1957: Numerical methods in weather predict ion: I1 Smoothing and f i l t e r i n g . Non. Wea. Rev., 85, 357-361.

“- -

Smith, iJ. L., H. M. Woolf, C. M. Hayden, and W. C . Shen, 1975: Nimbus-5 sounding data processing system part 11: Resul ts . Final Report , NASA Contract S-70249-A6, 35 pp.

S t ae l in , D. H . , A . H. Barrett , and J. VI. Waters, 1973: Microwave Spectrometer -

on the Nimbus 5 satel l i te : Meteorological and Geophysical data. Science, 1 8 2 , 1339. -

70

r

Wark, D. Q . , and D. T. Hilleary, 1969: Atmospheric temperature: Successful test of remote probing. Science, 165, 1256-1258.

Waters, J. W . , K. F. Kunzi, 3. L. Pettyjohn, R. K. L . Poon, and D . 11. Stae l in , 1975: Remote sensing of atmospheric temperature profiles with the Nimbus 5 microwave spectrometer. J . Atmos. Sci . , 32, 1953-1969. "-

Wilcox, R. W., and F. Sanders, 1976: Comparison of layer thickness as observed by Nimbus E microwave spectrometer and by radiosonde. J . Appl. Meteor., 15, 956-961.

"

-

71

1. REPORT NO. . 4 ~-

3. RECIPIENT'S CATALOG NO. -_

NASA RP-1070 a. T I T L E AND S U B T I T L E

"

] 5. REPORT DATE

~ ___ -


January 198 1 6. PERFORMING ORGANIZATION CCOE

Vance Moyer, and Nine-Min Cheng I 3. PERFORMING ORGANIZATION NAME AND ADDRESS 10. WORK UNIT, NO.

Department of Meteorology Texas A&M University

"327 1 1 . CONTRACT OR GRANT NO.

College Station, Texas 77843 13. TYPE OF REPORi ' 8: PERIOD COVERE1

~

2. SPONSORING AGENCY NAME AND ADDRESS

U.S . A r m y Research Office Reference Publication

Research Triangle Park, North Carolina 1.1. SPONSORING AGENCY CODE

~ ~~ I

15. SUPPLEMENTARY NOTES ~ -. - - - . ..

The u. S. Army Research Office has granted permission for NASA to publish these data for use in studies using space technology for weather-related programs.

16. ABSTRACT ~ ~~. - "_ -~ ..

This report addresses the question regarding the extent to which satellite sounding data can be used to determine atmospheric structure. Comparisons are made between rawinsonde and satellite profiles in seven areas for a wide range of surface and weather conditions. Variables considered consist of temperature, dewpoint temperature, thickness, precipitable water, lapse rate of temperature, stability, geopotential height, mixing ratio, wind direction, wind speed, and kinematic parameters, including vorticity and the advection of vorticity and temperature. In addition, comparisons are made in the form of cross sections and synoptic fields for selected variables.

data were linearly interpolated to obtain soundings coincident in time with the rawinsonde soundings. The TIROS-N data were obtained concurrently with the rawinsonde data, and no interpolation was performed. Results from the analysis of the discrepancies between satellite and rawinsonde data were similar for both types of satellite data. Biases were observed in both sets of satellite data as a function of altitude, and the discrepancies were approximately randomly distributed in the 1000-500, 500-300, and 300-100 mb layers.

flow patterns that agreed well with the rawinsonde wind fields. Surface wind patterns as well as magnitudes computed by use of the log law to extrapolate wind to a heighth of 10 m agreed well with observations.

data can be used to determine characteristics of large-scale systems but that small-

Sounding data frcm the NIMBUS-6 and TIROS-N satellites were used. The NIMBUS-6

Geostrophic wind computed from smoothed geopotential heights provided large-scale

The results of this study demonstrate rather conclusively that satellite profile

scale features, such as

frontal zones, caniq;.yet be resolved. 7. KEV WORDS DISTRIBUTION STATEMENT

Meteorological satellite data satellite-derived soundings satellite-derived winds AVE-SESAME data

Subject Category 47

t20. 3 . SECURITY CLASSIF. (Of t h h report) SECURITY CLASSIF . (oichi. WW)

P;. " N ~ E S Unclassified Unclassified I A05

22. PRICE ~

NASA-Langley, 1981

Documents

Tiros Satellite Profile Data - NASA · Tiros N Satellite Profile Data James R. Scoggins, William E. Carle, Keith Knight, Vance Moyer, and Nine-Min Cheng Texas AGM Uuivevsity College