Integrating multi-satellite and multi-model to estimate crop gross primary productivity at field level covering the U.S. Corn Belt

Kaiyu Guan, University of Illinois at Urbana-Champaign

Usage Details

Yaping Cai, Kaiyu Guan, Bin Peng, Bin Peng, Yunan Luo, Sibo Wang, Chongya Jiang, Wang Zhou, Tianyu Sun, Junrui Ni, Yizhi Huang, Kejie Zhao, Jingwen Zhang, Ziqi Qin, Ziyi Li, Sheng Wang, Yihong Jian, Xiyue Zhu, Ruike Zhu, Qu Zhou, Jack Nash, Maxwell Jong, Trevor Wong, Melissa Chen, Chen Song, Liling Chang, Lexuan Ye, Xiaocui Wu, Zhenrui Yue, Genghong Wu

Surface radiation data for shortwave and longwave at high spatial and temporal resolutions are highly demanded for both scientific research and societal applications. However, existing remote sensing products either focus on parts of surface radiation components or have low spatiotemporal resolutions; none of existing methodologies is able to estimate all-components surface radiation with high-spatial and high-temporal resolutions concurrently.

Here we propose to build a novel methodological/computational framework which integrates observations from new-generation geostationary and polar-orbit satellites, atmospheric radiative transfer models (ARTM), and land surface models (LSM) to provide accurate estimation of all components of surface radiation including both shortwave and longwave. The computational resource will be used to generate radiation component products, HyperRad, at high-spatial (500 m to  2 km) and high-temporal (5 minute) resolutions, under all-sky conditions, in near-real-time (1-day latency), and with spatiotemporally explicit quantitative uncertainties, covering the whole Continental United States (CONUS). The outcome products generated in this proposal will be very useful for ecosystem modeling, monitoring and mapping.

The computational demand of our satellite data interpretation and modeling effort is huge and the Blue Waters facility offers us the best solution for large processing element demands, high-frequency I/O, and output post-processing and visualization.