Impact of sub-pixel variations on ocean color remote sensing products.

Passive remote sensing of the Earth system has used spatial resolutions ranging from meters to kilometers. It is thus necessary to understand how data products with different spatial resolutions can be compared with each other, and how sub-pixel variations may affect data comparison. This is particularly important for ocean color remote sensing where the measured signal (water-leaving radiance or remote sensing reflectance) is a non-linear function of sub-surface constituents. As a result, products at low resolution are not necessarily arithmetic or geometric means of those at higher resolution. Here, we developed analytical expressions to link ocean color properties derived from high- and low-resolution data, and the proof-of-concept is further demonstrated with both simple examples and measurements of MERIS full-resolution (FR) and reduced resolution (RR). These results suggest that current global chlorophyll concentration is likely underestimated due to the coarse spatial resolutions. Application of the expressions will facilitate cross-sensor comparisons and may also reduce uncertainties.