复杂地形区短波及长波辐射量建模与反演研究

As a main driving force for the earth’s matter and energy cycle, land surface radiation budget is a key modulator of global climate change. The corresponding surface radiative flux components (e.g., net radiative fluxes) are also essential driving parameters for many models of land applications, ecosystem, hydrology and climate, etc. Although more and more attention has been paid to the derivation of surface radiation from space, most of the existing algorithms are only valid for horizontal surfaces without accounting for topographic effects over rugged terrain areas. As pointed out by many researchers, significant bias can be induced in the retrieved radiation if topographic effects are not considered, thus restricting the applications of the derived radiation product. Moreover, considering the large proportion of mountainous areas over the earth’s surface (2/3) and their important roles in climate change studies, modeling and retrieval of radiation over rugged terrain are highly imperative. This project is committed to building shortwave and longwave radiation operational inversion schemes over rugged terrain from space. First, a new integration algorithm that can simultaneously retrieve multi-component of shortwave and longwave radiation is dedicated to be developed based on Artificial Neural Network(ANN), that is, under the unified framework, the downward radiation, upward radiation and net radiation can be derived at the same time with the same manner, avoiding the atmospheric correction, calculation of surface temperature, surface albedo and thus avoiding multiple error transfer; Then, by fully quantifying topographic effect on radiation, such as, terrain shading, shadows, sky-view factors as well as emitted/reflected radiative fluxes from nearby terrain etc., operational shortwave and longwave topographic radiation models are intend to be built, so that get rid of the plight of lacks of operational model in this field. Finally, based on the terrain correction, all radiation components (shortwave and longwave) over rugged terrain can be directly derived by coupling the outputs of ANN and the proposed topographic radiation models.

地表辐射收支是影响全球气候变化的关键要素，同时也是气候、气象、生态及水文模型的重要驱动参数。当前辐射量遥感反演算法普遍假设地表为水平面，忽略地形影响，反演精度低，且产品难于被广泛应用，再加之全球陆表2/3被山区覆盖，复杂地形区辐射建模与反演研究势在必行。本项目致力于构建利用遥感估算复杂地形区短波及长波辐射量的整套反演体系。首先开发利用遥感数据直接估算水平地表短波及长波辐射量的新型一体化算法，即，在统一的反演框架下直接实现下行辐射、上行辐射以及净辐射的同时反演，避免大气校正、地表反照率和地表温度求解的多重误差传递和积累；考虑地形遮蔽、周围地形对目标像元的辐射贡献、阴影、天空可视比例等因素分别构建可用于大面积反演的复杂地形区短波及长波辐射模型，摆脱该领域缺乏可操作模型的困境。最后，在地形遮蔽校正基础上，将一体化反演的水平地表辐射量与复杂地形区辐射模型耦合实现复杂地形区短波及长波辐射量的反演。

地表辐射收支是地球系统能量、水分循环的重要驱动力，短波和长波辐射是气候、气象、生态、水文等应用模型的核心输入数据，对理解和认识全球变化有重要作用。受到云、气溶胶、地形等综合影响，当前利用遥感估算地表辐射收支仍有较大不确定性，所得遥感产品与实际应用需求仍有一定距离。本项目针对已有研究的不足，开展了两大方面的研究内容：（1）从多个角度开发了全天候短波辐射分量的遥感估算算法，实现了太阳总辐射、直射辐射和散射辐射的同步反演；光学与微波结合开发了全天候长波辐射的反演方法，同时，给出了全天候长波辐射的通用参数化方法；（2）考虑山区太阳遮蔽、天空漫反射、周围地形反射贡献等建立了复杂山区短波辐射模型（SWTRM），同时，考虑山区阴影、天空热辐射以及周围地形热辐射贡献等构建了山区长波辐射模型（LWTRM），这些模型的最大的优点是可以与现有平面地表遥感短波和长波辐射直接耦合，并得到DEM尺度的山区的辐射，是目前可用于大面积复杂山区遥感反演的辐射模型，该模型也在青藏高原地区进行了应用。.通过研究，从晴空条件到有云条件下的辐射，从短波辐射到长波辐射，再到复杂山区辐射建模和反演，本项目基本建立了全天候地表辐射遥感估算体系。项目基于开发的新型辐射反演算法和地形模型生产了青藏高原地区90m，1hr的太阳辐射数据，并已在山区冰川/积雪消融建模和水资源管理中得到成功应用。项目建立的全天候短波辐射、长波辐射遥感算法、复杂山区短波辐射模型和长波辐射模型将为辐射平衡遥感研究、全球变化研究等提供科学依据和新思路。