Hyperspectral Microwave Photonic Radiometer

Passive Microwave Remote Sensing is currently utilized by NASA, NOAA, and others to conduct Earth Science missions, including weather forecasting, early warning systems, and climate studies. These sensors could also be used in upcoming missions such as Saturn’s moon to study atmospheric makeup. On earth, humidity and temperature sounding is conducted near several absorption lines to determine key parameters of the atmospheric state, including moisture content, temperature, and barometric pressure. Using neural networks, these parameters are retrieved from raw sensor data at a small number of discreet frequencies near the absorption lines. To improve retrieval accuracy as well as predictive ability of weather models, a large number of closely spaced frequencies, i.e. hyperspectral sensing should be implemented. Such approach is also helpful in RFI mitigation and sensor calibration. As such, under the current program PSI will develop a hyperspectral microwave photonic radiometer concept capable of acquiring a hundred or more detection frequencies simultaneously on a ~1GHz channel spacing. We leverage microwave photonics to convert the RF signals up to optical frequencies, where they are readily manipulated with low loss and wide bandwidth using optical components. We survey available elements, such as prism, diffraction grating, or arrayed waveguide grating, to implement the dispersive function. Under a previous NOAA Phase II SBIR, PSI delivered a frequency agile microwave radiometer that could be tuned across the entire V-band with ~ 1K NETD at 2GHz bandwidth and 100msec integration time. This sensor, however, performed measurements in serial fashion by sweeping a local oscillator. Under the proposed effort, we will adapt our existing V-band hardware to perform these spectrally resolved measurements simultaneously. We will demonstrate an 8-channel system with 3 GHz frequency spacing, which can be scaled to larger channel counts in a straightforward manner.