Abstract
This paper provides methodologies developed and implemented by the NASA VIIRS Calibration Support Team (VCST) to evaluate the S-NPP VIIRS DNB on-orbit performance by monitoring the long term radiance and reflectance trends of DNB measurement over two stable ground sites. The DNB and M bands Sensor Data Records (SDRs) produced by the Interface Data Processing Segment (IDPS) and NASA Land Product Evaluation and Algorithm Testing Element (PEATE) are used in this study. The Libya 4 desert and Dome C snow surfaces sites are used to collect nearly nadir overpass data. Impacts of solar zenith angles on the observed radiance and reflectance were corrected. The radiance and reflectance long term trends are used to examine the DNB calibration stability and consistency. Simulated DNB radiance values from the aggregated M bands (M4, M5, and M7) are compared with the simultaneously-observed DNB radiance.
The results showed that since launch the DNB modulated RSR could have about 4% influence on its transmitted solar spectral irradiance. Meanwhile the modulated RSR impact on the M bands integrated solar spectral irradiance has been less than 0.065%. After excluding some early mission data points, the results derived using IDPS DNB data show a general upward trend (about 10% - 16%) in both radiance and reflectance. These trends are likely caused by inconsistent calibration used in the IDPS forward data processing, including the use of different SDSM screen transmission and algorithms, and RSR degradation used to compute and update the RSB gain coefficients. On the other hand, the IDPS M bands, including M4, M5, and M7 which are within the DNB wavelength range, did not show obvious upward trends. The theoretical reflectance long term trend calculated from SCIAMACHY Libya 4 surface spectra and DNB RSR data have indicated a decrease trend of about 1% in the past three years. By comparison, the long-term trends derived from the Land PEATE DNB and M bands data over Libya 4 and Dome C have been very stable.
Though we still need more VIIRS observations and thus longer time series to better understand and verify the IDPS and Land PEATE data quality, the results of this study have provided useful information for VIIRS post-launch calibration assessment and preliminary analysis of its calibration stability and consistency for the first 3 years.
Keywords: VIIRS, calibration, Day Night Bands, Reflective Solar Bands, TOA radiance, TOA reflectance
A Ground Reference Method For Suomi-NPP VIIRS Day-Night Band And M Bands Performance Validation
This paper provides methodologies developed and implemented by the NASA VIIRS Calibration Support Team (VCST) to evaluate the S-NPP VIIRS DNB on-orbit performance by monitoring the long term radiance and reflectance trends of DNB measurement over two stable ground sites. The DNB and M bands Sensor Data Records (SDRs) produced by the Interface Data Processing Segment (IDPS) and NASA Land Product Evaluation and Algorithm Testing Element (PEATE) are used in this study. The Libya 4 desert and Dome C snow surfaces sites are used to collect nearly nadir overpass data. Impacts of solar zenith angles on the observed radiance and reflectance were corrected. The radiance and reflectance long term trends are used to examine the DNB calibration stability and consistency. Simulated DNB radiance values from the aggregated M bands (M4, M5, and M7) are compared with the simultaneously-observed DNB radiance.
The results showed that since launch the DNB modulated RSR could have about 4% influence on its transmitted solar spectral irradiance. Meanwhile the modulated RSR impact on the M bands integrated solar spectral irradiance has been less than 0.065%. After excluding some early mission data points, the results derived using IDPS DNB data show a general upward trend (about 10% - 16%) in both radiance and reflectance. These trends are likely caused by inconsistent calibration used in the IDPS forward data processing, including the use of different SDSM screen transmission and algorithms, and RSR degradation used to compute and update the RSB gain coefficients. On the other hand, the IDPS M bands, including M4, M5, and M7 which are within the DNB wavelength range, did not show obvious upward trends. The theoretical reflectance long term trend calculated from SCIAMACHY Libya 4 surface spectra and DNB RSR data have indicated a decrease trend of about 1% in the past three years. By comparison, the long-term trends derived from the Land PEATE DNB and M bands data over Libya 4 and Dome C have been very stable.
Though we still need more VIIRS observations and thus longer time series to better understand and verify the IDPS and Land PEATE data quality, the results of this study have provided useful information for VIIRS post-launch calibration assessment and preliminary analysis of its calibration stability and consistency for the first 3 years.
Keywords: VIIRS, calibration, Day Night Bands, Reflective Solar Bands, TOA radiance, TOA reflectance