Long-term Solar Irradiance at the Surface Derived from Satellite Data

1
Long-term Solar Irradiance at the Surface Derived
from Satellite Data
Istvan Laszlo1(GOVERNMENT PRINCIPAL INVESTIGATOR)
, Hongqing Liu2 1NOAA/NESDIS/STAR/SMCD, 2DPSGS

• Requirements
• Describe and understand the state of the climate
  system through integrated observations, analysis,
  and data stewardship.
• Conducting observational, diagnostic, and
  modeling research to improve understanding of
  physical mechanisms and processes of climate
  variability and predictability that will lead to
  improved climate models and climate predictions.
• Science How does the solar radiation available
  at the surface respond to changes in atmospheric
  and surface conditions?
• Benefit NCEP and the climate modeling community
  will have accurate surface solar radiation
  products that provide observational constraints
  for climate models, and better quantify the
  effect of changes in the climate system.

Analysis of Surface Solar Radiation Trends
-Results
21-year Average Surface Solar Radiation
Data and Method

• Data source International Satellite Cloud
  Climatology Project (ISCCP) D1
• State of the art long-term radiance and cloud
  data since July 1983
• Spatial/temporal resolution 280-km equal area/3
  hours (nominal)
• Satellite source Operational polar and
  geostationary satellites including those of NOAA
• Algorithm Version 3 of the Satellite Algorithm
  for Shortwave Radiation Budget (SASRAB-V) is
  applied to the ISCCP D1 radiances for retrieving
  the shortwave (SW 0.2-4.0 m) irradiance at the
  surface.
• Retrievals are compared to two other,
  independently derived records of the SW surface
  irradiance variability of radiation with time is
  analyzed.

• The three datasets differ in the magnitude of
  trend and the number of years needed to detect a
  trend.
• Differences are larger in the 83-86 period.
• Order of anomalies are reversed at the beginning
  and at the end.
• The Pinatubo anomaly in GEWEX-SRB is the
  smallest.
• Trend (T) Wm-2/decade 95 CI 95 confidence
  intervals Yrs number of years needed to detect
  Trend with probability 0.9.
• Positive T TSASRAB gt TGEWEX gt TISCCP
• However, 95 CIs are large trends are not
  statistically significantly different.
• In some cases, the length of the record is not
  enough to detect the calculated trend.
• Equatorial (20S-20N) trends are larger than
  global ones.
• Hemispheric differences are the largest in
  GEWEX-SRB.
• NH trends in GEWEX and SASRAB are larger than SH
  it is the opposite in ISCCP (but large CI(!))

Main features of algorithms used for the retrieval of SW surface irradiance data Main features of algorithms used for the retrieval of SW surface irradiance data Main features of algorithms used for the retrieval of SW surface irradiance data Main features of algorithms used for the retrieval of SW surface irradiance data
SASRAB-V ISCCP-FD GEWEX-SRB
Radiances ISCCP-D1 ISCCP-D1 ISCCP-DX averaged to 1ox1o
Retrieval Delta-Eddington RT NASA GISS climate GCM RT LUT from delta-Eddington RT
Water vapor ISCCP-D1 NOAA TOVS GEOS-4
Ozone ISCCP-D1 TOMS TOMS
Aerosol Tegen et al. monthly climatology NASA GISS CM OPAC
SW surface irradiance (Wm-2) derived from the
ISCCP D1 data using the SASRAB-V algorithm are
averaged for each month for the years of
1984-2004.
Analysis of Surface Solar Radiation Trends -Method
Evaluation - Example over a high altitude station

• Normalized all fluxes to ISCCP-FD TOA downward
  flux
• Used deseasonalized mean monthly anomalies of SW
  fluxes for July 1983-June 2004 obtained averages
  for global (50S-50N), equatorial (20S-20N),
  hemispheric (50S-0S 0N-50N) regions
• Trend Analysis linear model (Weatherhead et
  al.1998)
• Used trends only for comparison ! Episodic
  events (Pinatubo) were not removed not the
  real trend.

Surface data Downward solar radiation data from
the GEWEX Asian Monsoon Experiment (GAME) Asian
Automatic weather station Network (AAN)
collection at Amdo, Tibet, China.
Science Challenges Retrieval of surface and TOA
fluxes and their components (direct and diffuse)
consistent with observations proper
representation of aerosol variability. Next
Steps Improve retrievals over polar regions
Calibrate SASRAB-V fluxes at TOA with
ERBE/CERES/GERB Transition Path Provide
long-term surface solar radiation data for the
follow-up to the GEWEX Radiative Flux Assessment.
Time series of satellite and ground fluxes at
Amdo. The grid-cell averages of satellite
estimates are similar and smaller than the ground
average.
Grass is the dominant vegetation at Amdo. The
ISCCP grid, however, contains complex surface
features grass, savanna, lakes and snow. Surface
elevation also varies significantly.
ISCCP grid cell 4979
Data Mean SD Min Max
Amdo 224 43 145 307
SASRAB-V 215 37 153 307
ISCCP-FD 211 41 137 308
Mean, standard deviation (SD), minimum (Min) and
maximum (Max) in Wm-2.