1
Wet and dry spells in the Rio Santa Basin, Tropical Andes of Peru Fabien Maussion, Maria Siller, Miriam Schlumpberger, Max Hartmann, and Wolfgang Gurgiser Department of Atmospheric and Cryospheric Sciences, University of Innsbruck, Austria [email protected] Context The Rio Santa basin in Peru (Ancash region): Strong west – east contrasts (dry wet) Rainy season: October to March, mostly (deep) convection Subsidence agriculture (rainfed) Vulnerable to climate variability and change Figure 1: Overview of the study area with a close up of the Rio Santa basin Motivation Precipitation variability strongly correlated to 200hPa zonal wind (Bolivian high) Conceptual framework formulated by Garreaud et al. (2003) for the Altiplano but never tested at such low latitudes Does the framework apply to the Rio Santa region as well? Is moisture availability really the only driver for wet & dry spells? Figure 2: Seasonally averaged (1979–2016) horizontal wind streamlines and geopotential height at 200hPa from ERA-Interim data Figure 3: Conceptual framework for rainy and dry episodes over the Altiplano (Garreaud et al., 2003) Definition of Wet and Dry spells Daily precipitation from 16 stations (1979–2016, variable quality) Wet spell: 50% of stations exceed the 70th percentile of daily rainfall amount for at least three consecutive days Dry spell: 50% of stations inferior to the 30th percentile of daily rainfall amount for at least five consecutive days Average of 3.3 wet spells per rainy season (duration 8 days) Average of 2.3 dry spells per rainy season (duration 10.8 days) Teleconnections: synoptic composite analysis (ERA-Interim) Figure 4: Composite analysis during wet and dry spells. Top: 200hPa winds and geopotential. Middle: 200hPa wind and temperature anomalies (95% confidence). Bottom: 500hPa wind composite and specific humidity anomalies. Figure 5: Cross section at 9.75°S. Top: wind. Bottom: wind and specific humidity anomalies. Case study: WRF simulations Figure 6: Selection of the simulation periods. Figure 7: WRF model domains. Figure 8: Simulated 3-day accumulated precipitation in the wet and dry cases (WRF 1km). Figure 9: Cross section at 9.75°S. Top: relative humidity. Bottom: zonal and meridional wind. Conclusions synoptic composite analysis roughly confirms previous conceptual framework – albeit coarse and qualitative detailed WRF modelling provides a much more nuanced picture ingredients for convection need to be studied in more detail! References: Garreaud, R. D., Vuille, M., and Clement, A. C. (2003). The climate of the Alti- plano: observed current conditions and mechanisms of past changes. Palaeo- geography, Palaeoclimatology, Palaeoecology Schlumpberger (2017) & Siller (2017), Master Theses: http://diglib.uibk.ac.at/urn:nbn:at:at-ubi:1-6985 http://diglib.uibk.ac.at/urn:nbn:at:at-ubi:1-7816 Acknowledgements: This study relies heavily on the two Master Theses listed above. The WRF simulations were run on the HPC infrastructure of the U. of Innsbruck.

Wet and dry spells in the Rio Santa Basin, Tropical Andes ... · Figure 5: Cross section at 9.75°S. Top: wind. Bottom: wind and specific humidity anomalies. Case study: WRF simulations

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Wet and dry spells in the Rio Santa Basin, Tropical Andes ... · Figure 5: Cross section at 9.75°S. Top: wind. Bottom: wind and specific humidity anomalies. Case study: WRF simulations

Wet and dry spells in the Rio Santa Basin, Tropical Andes of PeruFabien Maussion, Maria Siller, Miriam Schlumpberger, Max Hartmann, and Wolfgang GurgiserDepartment of Atmospheric and Cryospheric Sciences, University of Innsbruck, Austria – [email protected]

Context

The Rio Santa basin in Peru (Ancash region):• Strong west – east contrasts (dry→ wet)• Rainy season: ≈ October to March, mostly (deep) convection• Subsidence agriculture (rainfed)• → Vulnerable to climate variability and change

­80W ­70W ­60W ­50W ­40W

­20S

­10S

Eq.

10N

1000 km

RIO SANTABASIN

PACIFIC

ATLANTIC

VENEZUELA

COLOMBIA

EQU.

PERU BRAZIL

BOLIVIA

PARA­GUAY

AMAZONIANBASIN

ALTIP

LAN

O

­78.5W ­78.0W ­77.5W ­77.0W

­10.5S

­10.0S

­9.5S

­9.0S

­8.5S

100 km

HUARAZ

HUASCARAN6768 m

0

1000

2000

3000

4000

5000

Alti

tude

 (m

)

Figure 1: Overview of the study area with a close up of the Rio Santa basin

Motivation

• Precipitation variability strongly correlated to 200hPa zonal wind(Bolivian high)

• Conceptual framework formulated by Garreaud et al. (2003) forthe Altiplano but never tested at such low latitudes

• Does the framework apply to the Rio Santa region as well?• Is moisture availability really the only driver for wet & dry spells?

Figure 2: Seasonally averaged (1979–2016) horizontal wind streamlines and geopotential height at200hPa from ERA-Interim data

Figure 3: Conceptual framework for rainy and dry episodes over the Altiplano (Garreaud et al., 2003)

Definition of Wet and Dry spells

• Daily precipitation from 16 stations (1979–2016, variable quality)• Wet spell: 50% of stations exceed the 70th percentile of daily

rainfall amount for at least three consecutive days• Dry spell: 50% of stations inferior to the 30th percentile of daily

rainfall amount for at least five consecutive days• Average of 3.3 wet spells per rainy season (duration 8 days)• Average of 2.3 dry spells per rainy season (duration 10.8 days)

Teleconnections: synoptic composite analysis (ERA-Interim)

Figure 4: Composite analysis during wet and dry spells. Top: 200hPa winds and geopotential.Middle: 200hPa wind and temperature anomalies (95% confidence). Bottom: 500hPa windcomposite and specific humidity anomalies.

Figure 5: Cross section at 9.75°S. Top: wind. Bottom: wind and specific humidity anomalies.

Case study: WRF simulations

302010

0102030

ER

A-I

nte

rim

zonal w

ind s

peed

at

20

0 h

Pa (

m s−

1)

EASTERLIES

WESTERLIES

DRYcase

WETcase

0

2

4

6

8

10

GPM

pre

cipit

ati

on

(mm

day−

1)

P70

P30

Dec Jan Feb Mar Apr2014 2015 2015 2015 2015

stati

on

dry

/wet

spells

Figure 6: Selection of the simulation periods.

-80W -70W -60W -50W -40W

-20S

-10S

Eq.

10N D01 (WRF25)

D02 (WRF5)

D03 (WRF1)

Figure 7: WRF model domains.

­79W ­78W ­77W ­76W ­75W ­74W

­10S

­9S

WET pblACM

0

2

5

10

20

50

80

120

3 da

ys p

reci

pita

tion 

(mm

)

­79W ­78W ­77W ­76W ­75W ­74W

­10S

­9S

DRY pblACM

0

2

5

10

20

50

80

120

3 da

ys p

reci

pita

tion 

(mm

)

Figure 8: Simulated 3-day accumulated precipitation in the wet and dry cases (WRF 1km).

78 77 76 75longitude (deg W)

0

1

2

3

4

5

6

7

alt

itude (

km a

sl)

Huaraz

EP AB

WET: Day 1, 15:00 LT

75

80

85

90

95

rela

tive h

um

idit

y (

%)

78 77 76 75longitude (deg W)

0

1

2

3

4

5

6

7

alt

itude (

km a

sl)

Huaraz

EP AB

DRY: Day 2, 15:00 LT

75

80

85

90

95

rela

tive h

um

idit

y (

%)

78 77 76 75longitude (deg W)

0

1

2

3

4

5

6

7

alt

itude (

km a

sl)

Huaraz

EP AB

quiverbckg5 m s−1WET: Day 1, 15:00 LT

8

4

0

4

8

meri

dio

nal w

ind s

peed (

m s−

1)

78 77 76 75longitude (deg W)

0

1

2

3

4

5

6

7

alt

itude (

km a

sl)

Huaraz

EP AB

quiverbckg5 m s−1DRY: Day 2, 15:00 LT

8

4

0

4

8

meri

dio

nal w

ind s

peed (

m s−

1)

Figure 9: Cross section at 9.75°S. Top: relative humidity. Bottom: zonal and meridional wind.

Conclusions

• synoptic composite analysis roughly confirms previousconceptual framework – albeit coarse and qualitative

• detailed WRF modelling provides a much more nuanced picture• ingredients for convection need to be studied in more detail!

References:Garreaud, R. D., Vuille, M., and Clement, A. C. (2003). The climate of the Alti-plano: observed current conditions and mechanisms of past changes. Palaeo-geography, Palaeoclimatology, PalaeoecologySchlumpberger (2017) & Siller (2017), Master Theses:http://diglib.uibk.ac.at/urn:nbn:at:at-ubi:1-6985

http://diglib.uibk.ac.at/urn:nbn:at:at-ubi:1-7816

Acknowledgements:This study relies heavily on the two Master Theses listed above.The WRF simulations were run on the HPC infrastructure of the U. of Innsbruck.