CoLM改進與多源遙感陸面數據同化系統研發

馬明國++韓旭軍++黃春林++鄭小谷++上官微

摘 要：該報告以發展氣候系統模式為核心，研發相關的模擬技術，確定關鍵物理和模發展的冠層輻射模型為三維輻射傳輸模型（簡稱3-D模型）以替換陸面模型CoLM中基于二流近似的輻射傳輸模塊，將動力植被模塊（DGVM）引入了陸面模型CoLM，實現了模型的改進；制備了2005—2010年比濕、氣壓、氣溫、風速（U，V）、降雨、入射短波輻射、入射長波輻射大氣驅動數據集和土壤及植被功能型參數數據集；實現了一個高分辨率多源遙感陸面數據同化系統HDAS，包含了多種數據同化算法，并針對高分辨率陸面數據同化需要重點解決了高性能計算問題；能夠在陸面過程模型的動力學框架內同化多源遙感觀測以及其他觀測資料，生產中國陸地區域內2005—2010年空間分辨率為5 km、時間分辨率為1 h的同化數據集。

關鍵詞：通用陸面模式 CoLM 動態植被 陸面數據同化 多源遙感

Abstract： The objective of this project is to develop the climate modelling techniques， in which the new 3D canopy radiative transfer model will be used to replace the two-stream approximations for radiative transfer of CoLM； the dynamic global vegetation model （DGVM） was implemented in CoLM. The hourly atmospheric forcing data between 2005 and 2010 were produced for relative humidity， air pressure， air temperature， wind speed （U，V）， precipitation， incident shortwave radiation， incident longwave radiation. The soil properties and plant functional type properties were also prepared. A high resolution land data assimilation system （HDAS） for multi-sources remote sensing data was developed. HDAS includes several kinds of data assimilation algorithms， and was designed for high performance computing. The multi-sources remote sensing data can be merged into land surface model （CoLM） within HDAS to produce the land surface data sets with 5 km spatial resolution and 1 hour temporal resolution between 2005 and 2010.

Key Words： Common land model； CoLM； DGVM； Land data assimilation system； Multi-sources remote sensing

閱讀全文鏈接（需實名注冊）：http：//www.nstrs.cn/xiangxiBG.aspx？id=12642&flag=1