New Capabilities of Diffuser Calibration Lab at GSFC NASA to Support Remote Sensing

Yaokai Liu,Ning Wang,Xinhong Wang,Yonggang Qian

2016-01-01

Abstract:The USGS ROLO model predicts the top-of-the atmosphere lunar irradiance as a function of lunar phase and libration angles over the spectral range from 350 nm to 2350 nm. As demonstrated with SeaWiFS, by measuring the lunar irradiance on a routine basis and comparing the measurements with the ROLO model, the relative trending of a satellite sensor band’s responsivity can be achieved with uncertainties as low as 0.13 %. However, the model requires further measurements to be made before the moon can be used as an exoatmospheric calibration source with an absolute scale and SI traceability. The current scale uncertainty in the absolute lunar irradiance predicted by the ROLO model are estimated to be on the order of 5-10 %. Recently, two ground-based absolute SI-traceable top-of-the atmosphere measurements of the lunar irradiance were made over the spectral range from 380 nm to 1050 nm, at phase angles of 7o and 17o, with uncertainties on the order of 0.5 %. Implications of tying the absolute lunar irradiance scale to low uncertainty, TOA lunar measurements are discussed. If tied to the absolute SI-traceable measurements, the output of the ROLO model is itself SI-traceable with quantified uncertainties an order of magnitude less than the estimated absolute uncertainties in the current ROLO model. Additional empirical adjustments to the phase and libration angle dependence of the output of the ROLO model using lunar observation data from Earth remote sensing satellite sensors are presented. Implications of an empirically revised ROLO Model output for transitioning the Moon toward an absolute, exo-atmospheric, SI-traceable satellite sensor calibration source with uncertainties commensurate with CLARREO science data product requirements will be discussed. Characterize VIIRS DNB Calibration Coefficients to Improve its Radiometric Performance Sirish Uprety – Colorado State University/ Cooperative Institute for Research in the Atmosphere (CIRA); Changyong Cao – NOAA/NESDIS/STAR; Xi Shao – University of Maryland ABSTRACT: S-NPP VIIRS day/night band uses three gain stages. Gain for low gain stage (LGS) is computed using solar diffuser which is similar to other RSB bands. However, for medium and high gain stages, gain ratio is calculated once a month using DNB observations near terminator crossing. Gain ratio values are multiplied with LGS gain to obtain the gains for MGS and HGS. Offsets for all gain stages are computed using DNB observations over Pacific Ocean. Both of the above steps are performed during new moon. Offset value is computed, one for each detector over each sample of an aggregation zone. However, for gain ratio, one value is computed for each aggregation zone. S-NPP VIIRS day/night band uses three gain stages. Gain for low gain stage (LGS) is computed using solar diffuser which is similar to other RSB bands. However, for medium and high gain stages, gain ratio is calculated once a month using DNB observations near terminator crossing. Gain ratio values are multiplied with LGS gain to obtain the gains for MGS and HGS. Offsets for all gain stages are computed using DNB observations over Pacific Ocean. Both of the above steps are performed during new moon. Offset value is computed, one for each detector over each sample of an aggregation zone. However, for gain ratio, one value is computed for each aggregation zone. The assumption is that the gain ratio remains nearly constant over an aggregation zone. This study focuses on analyzing the gain ratio over each aggregation zone, the differences observed for samples within an aggregation zone and its impact on DNB SDR product. It was observed that samples for aggregation zones near the edge of the scan exhibit high variability in gain ratio. There are cases when the gain varies by more than 15% within an aggregation zone for a detector. The study indicates that computing gain ratio on sample by sample basis over each aggregation zone can help improve the radiometric accuracy of the DNB SDR product especially for higher aggregation zones towards the edge of scan. In addition, the study can also be helpful to characterize and improve future J1 VIIRS DNB which has suggested significant non-linearity, especially for HGS, near the edge of the scan.