IceCube Level 1 Data Calibration and Analysis

The IceCube mission is the latest NASA's effort to advance the technology readiness (TRL) of the 883 GHz commercial radiometer on a 3U cubesat. Exceeding the designed lifetime of ~ 1 month and original expectation of technology demonstration, IceCube orbited the Earth for more than 15 months, producing the first ever global cloud ice mass map. The IceCube 883-GHz observations remain the only global survey of cloud-induced scattering radiances at this frequency, and the statistical properties of these cloud signatures have a unique value to cloud remote sensing and climate science research.

In this presentation, the 15-month of calibrated IceCube Level 1 radiance data will be presented. IceCube is an essentially free-running radiometer without onboard calibration targets. It relies on an onboard switching circuitry to calibrate its backend electronics. The end-to-end radiometric calibration is based on periodical space views and the modeled Earth atmospheric clear-sky radiances. Because cloud-induced radiance (Tcir) is the difference between the observed and modeled clear-sky radiances, a small bias in the modeled clear-sky radiances has a minimal impact on Tcir. In this study we will demonstrate an innovative method for IceCube calibration, which uses machine learning/artificial intelligence (ML/AI) to capture the count variations during cold-space observations. For a free-running radiometer, the instrument gain and cold-space background are a strong function of time from power-on, ambient operation temperature, as well as cubesat spin rate. The ML/AL method is able to successfully reproduce the large radiometer count variability and suppress the cold-space measurement noise to ~ 2 counts (1 K in brightness temperature). The gain model is updated on a self-aware timeframe to account for the orbital change and receiver degradation during IceCube's lifetime. Using collocated CloudSat-IIR-IceCube measurements, an empirical ice water path (IWP) retrieval algorithm is developed, and the first global 883 GHz IWP distribution and statistics will be presented in conjunction with other microwave and infrared measurements of frozen particles.