You are not logged-in. Registered users may open a session for accessing more functions.
 If your are a registered user, you will access more function when logged-in here. 
 Go to the Home page   Products  
 Low Earth Orbiter Sea Surface Temperature   Sea Surface Temperature on Atlantic excepted High Latitudes   Surface Solar Irradiance on Atlantic excepted High Latitudes   Downward Longwave Irradiance on Atlantic excepted High Latitudes  
 Surface Solar Irradiance on Atlantic excepted High Latitudes(Images)   Surface Solar Irradiance on Atlantic excepted High Latitudes(Validation)   Surface Solar Irradiance on Atlantic excepted High Latitudes(Climatology)   Surface Solar Irradiance on Atlantic excepted High Latitudes(Presentation)  

LML Surface Solar Irradiance
Validation

SSI Match to Data Base

The calculated SSI is routinely compared to pyranometer measurements over land to estimate its accuracy, using a Match Up Data Base (MDB), which contains coincident pyranometer measurements and calculated SSI :

  • calculated SSI : average on a 0.3 degree square box centered on the pyranometer station.
  • measured SSI : hourly value centered at the time of the satellite image, obtained by integration of several measures (or interpolation).

    • The MDB is updated in near real time for satellite data and off line for pyranometer measurements, with a delay of about 50 days. Non automatic updates may occur in cases of numerous missing measurements.

    The validation is made on a limited set of measurement stations :

  • List of pyranometer stations.
  • Map of pyranometer stations : An abbreviated name is indicated for each station and different colors are used for the various data networks.

    SSI validation results

    Error statistics (where error means calculated SSI minus measured SSI) are calculated monthly, separately for each satellite, GOES and MSG, and globally on the LML area. The errors are expressed in W/m2 and in percentage of the mean mesure. The last available results are presented :

    A crude filtering is applied to the monthly results, however it can only eliminate the cases where one station is strongly erroneous compared to the others.
    A more careful validation has been made on one year data set, but restricted to the U.S.A. stations, it had been presented in [conference paper].