Global retrieval of harmonized microwave land surface emissivity leveraging multi-sensor measurements from GMI, AMSR2 and MWRIs

Accurate accounting of microwave land surface emissivity (MLSE) facilitates applications associated with monitoring the ecohydrological dynamics in global terrestrial ecosystems, quantifying cross-sphere carbon and water exchanges, and meeting the critical accuracy required for assimilation in the global precipitation retrieval algorithm. However, collaborative applications of emissivity retrieved from individual sensors are severely hampered by discrepancies in retrieval techniques, instrumental configurations, calibration errors, and broken temporal periods. To mitigate this gap, we present an innovative framework to retrieve harmonized emissivity from observations of five passive microwave sensors, namely, GPM-CO/GMI, Fengyun-3B, −3C and −3D/MWRI, and GCOM-W1/AMSR2. Six geostationary visible and infrared imagers onboard three geostationary platforms, i.e. GOES-16/ABI, Himawari-8, −9/AHI, and MSG-1, −2, −3/SEVIRI, were collocated to jointly provide clear-sky masks covering the globe. The simultaneous conical overpass (SCO) recalibration technique was applied to scale all emissivity subsets retrieved from different sensors to be aligned with the GMI retrievals across various land types. Quantitative analyses reveal exceptionally strong consistency among emissivities across different subsets (Pearson R ≈ 0.95, RMSD <0.011, and mean bias within ±0.005). Our estimates at 10.65 GHz show strong agreement with in-situ radiometer measurements over two grass and crop fields, with errors generally within ±0.01 at vertical polarization and a systematic underestimation of approximately −0.02 at horizontal polarization. Globally, we evaluate the recalibrated emissivities against four reference datasets derived using various techniques, which includes three single-sensor retrievals from GMI and AMSR-E, as well as a climatology emissivity atlas generated using the Tool to Estimate Land Surface Emissivity at Microwaves and Millimeter waves (TELSEM). The results demonstrate strong consistencies at both vertical ( R = 0.8–0.9, RMSD <0.015 or ∼ 1.5 %) and horizontal ( R = 0.9–0.95, RMSD <0.02 or ∼ 2 %) polarizations on a monthly scale. The observed discrepancies are primarily attributed to differences in instrumental configurations, calibration accuracy, and retrieval methodologies. The harmonized retrieval algorithm and the sophisticated cross sensor calibrations facilitate its implementation as a self-consistent emissivity data for various applications associated with terrestrial ecohydrological dynamics, surface hydrological properties estimation, as well as the physical-based precipitation retrieval algorithms over land. • Community retrievals of surface emissivity from five passive microwave sensors. • Cloud removals using joint observations from six geostationary VIS/IR imagers. • Sophisticated intercalibrations harmonizing emissivity subsets of different sensors.