Reflectance based vicarious calibration of HySIS sensors and spectral stability study over pseudo-invariant sites

Recently launched ISRO's Hyperspectral Imaging Satellite (HySIS) carrying two sensors, VNIR (Visible and Near Infrared) and SWIR (Short Wave Infrared) covering spectral region 420 nm to 2500 nm with better than 10 nm bandwidth and providing 30 m spatial resolution covering 30 km swath. During initial phase operation, in order to verify the performance of HySIS sensors carried out vicarious calibration coinciding with satellite over pass. From 12
^th
to 15
^th
December, 2018 in-situ measurements were carried out using field based instruments (ASD, Sunphotometer and Ozonometer) covering a portion of Karnataka. Field collected radiance and reflectance data for large homogeneous features (water, vegetation, soil and rock) are used to arrive to at calibration coefficients for DN to radiance conversion. During mission life, to study spectral stability of the sensor with in-situ measurements can be difficult, labor intensive and costly. Alternate is to use pseudo-invariant test sites which do not need simultaneous in-situ measurements because the rate of change is slow enough. These sites are well characterized and are used routinely by many satellite missions for cross sensor calibration. Hence, periodic acquisition of HySIS sensor is planned over the pseudo-invariant test sites. Accordingly, HySIS data is acquired over three sites, Ivanpah Playa (USA), White Sands (USA) and Gobabeb (Namibia) during summer month. Surface reflectance is generated using 6S RT code along with necessary atmospheric and sensor parameters covering Ivanpah Playa site. HySIS reflectance over playa surface shows spectrally flat, similar to space-borne EO-1 Hyperion and airborne AVIRIS hyperspectral sensors. However, variations in magnitude across three data sets acquired in same season were observed. In the case of HySIS reflectance in the spectral region 440 to 700 nm is slightly higher compared with Hyperion, however, maximum difference is less than 5%. Likewise, in the wavelengths beyond 2000 nm reflectance in HySIS is slightly lower compared with Hyperion, however, maximum difference is less than 5%. Similar variations in magnitude were observed when compared with AVIRIS. However, the differences in retrieved reflectance's can be due to variation in the atmospheric parameters, although data is acquired during same season.