Upload
prisca
View
50
Download
0
Embed Size (px)
DESCRIPTION
Micro-Pulse Lidar Network (MPLNET): 10 Years of Trying to Imitate AERONET. Principal Investigator: Judd Welton, NASA GSFC Code 613.1 Instrumentation & Network Management: Sebastian Stewart, SSAI GSFC Code 613.1 Data Processing & Analysis: Larry Belcher, UMBC GSFC Code 613.1 - PowerPoint PPT Presentation
Citation preview
Micro-Pulse Lidar Network (MPLNET): 10 Years of Trying to Imitate AERONET
Principal Investigator:Judd Welton, NASA GSFC Code 613.1
Instrumentation & Network Management:Sebastian Stewart, SSAI GSFC Code 613.1
Data Processing & Analysis:Larry Belcher, UMBC GSFC Code 613.1Tim Berkoff, UMBC GSFC Code 613.1
James Campbell, UCAR/Naval Research Lab
Administrative Support:Sonia Cyrus, SSAI GSFC Code 613.1
GLAS Validation Activities:Jim Spinhirne, formerly NASA GSFC Code 613.1
Judd Welton, Tim Berkoff
CALIPSO Validation Activities:Judd Welton, Tim Berkoff, James Campbell
AERONET & Synergy Tool Partnership:Brent Holben, NASA GSFC Code 614.4Dave Giles, NASA GSFC Code 614.4
NASA SMART-COMMIT Field Deployments:Si-Chee Tsay, NASA GSFC Code 613.2
Jack Ji, UMCP GSFC Code 613.2
Site Operations & Science Investigations…. many network partners around the world
MPLNET is funded by the NASA Radiation SciencesProgram and the Earth Observing System
MPLNET information and results shown here are theresult of efforts by all of our network partners!
May 2, 2001 May 3, 2001
00:00 12:00 00:00 12:00 00:00Morning AfternoonNighttime Nighttime Morning Afternoon
Time UTC
Alti
tude
(km
)
Tropospheric Aerosol from Asia
Well Mixed PBL Well Mixed PBLStratified PBL Stratified PBL
Asian aerosol entrainedwithin boundary layer
PBL Growth PBL Decay
Level 1 MPLNET Signals from NASA Goddard
Micro-Pulse Lidar Network (MPLNET)MPLNET Data Products
MPLNET Data Products:
Level 1 NRB Signals, Diagnostics(near real time, no quality screening)
Level 1.5 Level 1.5b: Aerosol, Cloud, PBL Heights and Vertical Feature MaskLevel 1.5a: Aerosol Backscatter, Extinction, Optical Depth Profiles and Lidar Ratio(near real time, no quality screening)
Level 2 Operational Products Under Development (beta data available upon request)(not real time, quality assured)
All data are publicly available in netcdf format. Errors included for all data products.
Data policy same as AERONET. We are a federated network, individual site providers deserve credit.
near real time: 1 hour or 1 day
MPLNET Data Products: Cirrus Detection for AERONET AOD Evaluation
Susceptibility of Aerosol Optical Thickness Retrievals to Cirrus Contamination during the BASE-ASIA CampaignJ. Huang, N.C. Hsu, S.C. Tsay, M.-J. Jeong, B. Holben, E.J. Welton
Micro-Pulse Lidar Network (MPLNET)MPLNET Processing
€
PNRB(r) = C βM (r) + β P (r)( )e−2 σ M ( ′ r )d ′ r 0
z∫ e−2 σ P ( ′ r )d ′ r 0z∫
Level 0, Raw Lidar Signal (counts/time):
Level 1 Signal:
Level 1.5 and 2.0 Products:
€
PMPL(r) =O(r)E
Dr2Cβ (r)e−2 σ ( ′ r )d ′ r 0
z∫ ⎧ ⎨ ⎩
⎫ ⎬ ⎭+
A(r)
D+
N s+ Nd
D
AfterpulseLaser – Detector Crosstalk
(unique to transceiver design)
Dark Counts
Overlap (up to 4 – 6 km,Long due to transceiver design)
Calibration “Constant”(range only, varies w/ time)Laser
PulseEnergy
Detector Deadtime
€
C , Layer Heights/Classification
βP (r) , σ P (r) , Lidar Ratio ˜ σ P˜ β P
⎛
⎝ ⎜
⎞
⎠ ⎟ , τ P col , τ P pbl
Deadtime provided by detector manufacturer. Considered fixed for lifetime of detector.
Laser energy and standard deviation measured in real time and stored in raw data file.
Solar background counts obtained from data between 45 – 55 km effective range, stored in raw data file.
The optical alignment of the MPL is checked (and fixed) in lab prior to deployment using our collimator (far field transmit/receive path). Ideal temperature of instrument recorded.
The following calibrations must be performed periodically and are rigorously evaluated:
Dark count (initial calibration, then monthly)Afterpulse (initial calibration, then monthly)Overlap (initial calibration, evaluated several times per year)Calibration constant (generated daily from AERONET, evaluated throughout year)
MPLNET Level 0 Calibrations
Dark Count:Lid on telescope, Laser energy off, record data for ~10 min
Afterpulse:Lid on telescope, Laser energy at nominal setting, record data for ~10 min
MPLNET Level 0 Calibrations: Dark Count & Afterpulse
Total Signal
Afterpulse
Darkcount
MPLNET Level 0 Calibrations: Overlap & Calibration Constant
The telescope temperature recorded during the calibration becomes the “set-point” temperature for operations.
Deviations from this value change the overlap shape.
AERONET AODused to calculatecalibration constantin molecular zone
We are testing athermal telescopes as asolution to the temperature-overlap problem
MPLNET Level 1 Data: Operations
MPL Maintained by on-site personnel
Must follow MPLNET specifications/protocol
Climate controlled environment, goal to achieve +- 2° C centered on set point temp
Some sites have special window wash and shutter mechanisms (maintain window quality, shade from sun in tropics at noon)
Afterpulse and darkcount calibrations performed monthly
Data transferred to NASA GSFC daily/hourly via FTP. Push or Pull.
Higher level products are retrieved daily and monitored weekly.
Q/C performed at Level 2
MPLNET Level 1 Data: Operations
MPLNET Level 1 Data: Operations
MPLNET Level 2 Data: Quality Control
MPLNET Level 2 Data: Manual Assessment of Data Screens
MPLNET Level 2 Data: Manual Assessment of Data Screens
MPLNET Level 2 Data: Manual Assessment of Data Screens
MPLNET Level 2 Data: Manual Assessment of Data Screens
Data that are qualified are stored as L2, sunphotometer constrained results
The calibration values from these data are interpolated and used to generate a continuous calibration function for the assessment period. These are used to generate L2 the continuous, gridded aerosol products.
MPLNET Level 2 Data: Sample Results
MPLNET Level 2 Data: Sample Results
MPLNET Level 2 Data: Sample Results
MPLNET Level 2 Data: Sample Results
Validation of MPLNET Aerosol Products: numerous MPL comparisons with AATS since 1997
How well do state-of-the-art techniques measuring the vertical profile of tropospheric aerosol extinction compare?
B. Schmid, R. Ferrare, C. Flynn, R. Elleman, D. Covert, A. Strawa, E. Welton, D. Turner, H. Jonsson, J. Redemann, J. Eilers, K. Ricci, A. G. Hallar, M. Clayton, J. Michalsky, A. Smirnov, B. Holben, and J. Barnard, J. Geophys. Res., 111, D05S07, doi:10.1029/2005JD005837, 2006.
Aerosol Extinction & Optical Depth profiles compared in comprehensive study:• MPLNET (column AOT anchored to AERONET)• MPL from ARM• Airborne Ames Sunphotometer (AATS) • Airborne in-situ (nephelometer & absorption photometer)• Airborne cavity ring-down system (Cadenza)• Ground-based Raman lidar (CARL)
AATS used as truth• AATS is most direct measure of AOT profile• AATS Column AOT within 2% of AERONET
AOT
Alti
tude
(km
)
Extinction 1/km
Alti
tude
(km
)
MPLNETMPL ARMNeph/PSAPAATS
Example Comparisons from study (25 Days total).Also surveyed past 7 field campaigns of such work.
Instrument (visible wavelengths, ~520 nm) Bias Error relati ve to AATS Cadenza - 13% Neph / PSAP - 15% MPLNET + 13% MPL ARM + 24% CARL (Ra man) + 54% Neph / PSAP Mean from 7 Ca mpa igns - 17% Size Dist. Der ived Mean from 3 Campa igns - 18% MPLNET Mean from 5 Ca mpaigns < + 20% ( work in prog ress)
Summary of Study Results: MPLNET among best performers
Conclusion: state-of-the-art techniques remain 15 - 20% uncertain
MPLNET meets or exceeds that target range
AATS-14 shown above
The Micro Pulse Lidar Network (MPLNET): Validation
Tim Berkoff is working on Lunar AOD capability with Cimel sunphotometer(examine our nighttime retrievals)
Would like to initiate lidar comparisons in theWashington DC – Baltimore region