Our proposal responds to the access technology focus area 2.1.1.2: creation of training data. The purpose of this project is to recast the 32-year record of satellite observations of the earth s…

Remote Sensing Systems (RSS), a Santa Rosa, California-based satellite data processing firm, received a $1.45 million Cooperative Agreement from NASA's Shared Services Center, with performance completion targeted for December 31, 2024. The contract addresses the Access Technology Focus Area 2.1.1.2 requirement for creation of training data by converting a 32-year historical record of satellite microwave observations into datasets compatible with machine learning applications. The project tackles critical technical barriers that have historically limited participation from the broader machine learning community in Earth science research by preparing microwave radiance data for accessibility and usability in advanced computational models.

The initiative focuses on standardizing microwave observations collected across disparate satellite instruments with varying measurement frequencies and observation angles, addressing the fundamental challenge of harmonizing data with differing technical specifications. RSS will perform specialized data processing work to resolve issues related to swath-form data delivery and irregular measurement footprint spacing, ultimately enabling collocation of microwave measurements with other Earth observation data and geophysical surface characteristics. This effort directly supports broader climate and environmental monitoring objectives by facilitating machine learning algorithm development for retrieving critical geophysical parameters including sea ice extent, ocean winds and temperature, atmospheric moisture, and vegetation cover. The contract contains no set-aside designation, reflecting open competition for this specialized research effort.

Our proposal responds to the access technology focus area 2.1.1.2: creation of training data. The purpose of this project is to recast the 32-year record of satellite observations of the earth s natural microwave emission into datasets that are compatible with machine learning. The microwave spectrum contains a wealth of information on our ever-changing planet. From providing information on sea ICE, to ocean winds and temperature, to atmospheric moisture, to soil and vegetation cover, the microwave spectrum is unique and vital to understanding earth. Records of these geophysical parameters can all be calculated using a combination of observed microwave radiances from satellites and relevant ancillary geophysical parameters associated with their corresponding earth scenes. However, to assemble and use the entire record of satellite microwave observations, measurements from a number of disparate instrument types need to be combined. These instruments often differ in seemingly arcane measurement details, such as the exact frequency of the measurement and the angle at which the earth is observed. This presents a challenge as measurement differences lead to differences in the observed radiances that need to be characterized and removed before the data can be combined or used in a common retrieval algorithm. Another challenge is that the radiance data are typically available in swath form, with elliptical measurement footprints that are irregularly spaced in latitude and longitude. This makes IT difficult to collocate the measurements taken by microwave imaging radiometers with other measurements, or with known characteristics of the earth s surface. Because of these challenges, algorithm development has generally been undertaken by remote sensing specialists, instead of attracting the attention of the large and growing machine learning community. The purpose of the work proposed here is to perform the specialized work that currently serves as a barrier to more widespread study of microwave measurements. This will lead to the creation of an easy-to-use datasets which will subsequently be made available to researchers and other users worldwide. We propose to perform the following for a large collection of microwave radiometers: adjust the measurements from the disparate instruments so that they are referred to a common set of frequencies at a single measurement angle. Resample the measurements (using a minimal-noise technique) onto a regular, latitude and longitude earth grid. Make available (on an identical grid) a set collocated ancillary variables that we deem likely to aid algorithm development. Examples of these data are: (II) meteorological information, including atmospheric profiles of temperature, moisture, the presence rain, surface temperature, and wind. (II) information about earth s surface, such as open water fraction and land use/land cover. (III) informative quality flags to aid the user in culling measurements that are inappropriate for their purpose. Retrieve surface emissivity for each frequency using atmospheric parameters from reanalysis on an identical regular earth grid. Use these emissivities and forward microwave models of the atmosphere to produce a second set of training data constructed using diverse, exactly specified atmospheric conditions. All of this would be freely provided in easy-to-use formats via popular web interfaces to maximize the potential for widespread use. Taken all together, we would provide 150 satellite-years of data that, by the end of the proposed work, would span more than 35 years of recent earth history.