Cross-calibration of Irs-1c, -1d and -p6: Wide Angle Sensors Using Synchronous or near Synchronous Matching Scenes

For many applications (e.g., disaster management, precision farming), it is required to monitor events using multi-date satellite imagery acquired over the same area. This requirement can be best met with a set of satellites whose radiometric characteristics are fully modelled. Even though there is a gamut of satellites with identical or near-identical radiometric characteristics, systematic difference between their measured radiance exists due to inherent variation in sensor response and viewing geometries. Inter-sensor calibration exercises are carried out to establish radiometric relationship between the sensors of similar spectral bands so that their products can be used in conjunction with the other for deriving the change in events. In this paper, we describe method followed and results of the inter-calibration of WIde Field Sensors (WIFS) of IRS-1C and -1D spacecraft and Advanced WIFS (AWIFS) of IRSP6 spacecraft sensors using near synchronous acquisitions over the Thar desert, serving in this study as a stable natural target. The WIFS payload has 2 spectral bands, B3 (0.62 – 0.68 μm) and B4 (0.77 – 0.86 μm) at spatial (ground) resolution of 188m with 5 days revisit time. IRS-P6 AWIFS payload has 4 spectral bands, B2 (0.52 – 0.59 μm), B3 (0.62 – 0.68 μm), B4 (0.77 – 0.86 μm) and B5 (1.55 – 1.70 μm) at spatial (ground) resolution of 56m with 5 days revisit time. In this study, it is assumed that the surface and atmospheric properties are relatively unchanged between the two data acquisitions. Using pre-launch coefficients, the top-ofatmosphere radiance of the target was computed from these data sets. Intrinsic variation present in their spectral response curves of each sensor was normalized band-wise with the pre-launch measured data. Multiple data sets acquired at different dates were used to evaluate this coefficient to make it usable for all the seasons. A new set of data was taken for verifying accuracies of the inter-sensor coefficients obtained from this study. A good correlation was found to exist among the data sets of the sensors after the inter-sensor calibration methodology was put to use.