We are developing methods to generate estimates of unmeasured spectra using other relevant data. These techniques generate data of interest to Earth scientists, thereby yielding a greater return on the investment in collecting and storing a vast amount of Earth science data.
Many applied science questions that are relevant to the earth science remote sensing community require analysis of enormous amounts of data that were generated by instruments with disparate measurement capabilities. Virtual Sensors is a method that uses models trained on spectrally rich (high spectral resolution) data to "fill in" unmeasured spectral channels in spectrally poor (low spectral resolution) data, thereby extracting more information from spectrally poor data.
We will analyze more images from a greater variety of instruments to solve a variety of earth science problems. We will develop more scalable algorithms to enable rapid generation of data products of interest to earth scientists.
( For details, see our Publications page)
We demonstrate our method by using models trained on the high spectral resolution Terra Moderate Resolution Imaging Spectroradiometer (MODIS) instrument to estimate what the equivalent of the MODIS 1.6- µm channel would be for the National Oceanic and Atmospheric Administration Advanced Very High Resolution Radiometer (AVHRR/2) instrument. Simulation of the 1.6- µm channel improves the ability of the AVHRR/2 sensor to detect clouds over snow and ice, enabling such detection further back in time than what is possible so far.
Ashok Srivastava, Ph.D.
Nikunj C. Oza, Ph.D. (co-I)
Virtual Sensors for Earth Science Paper - Broken