A statistical model for the selection of ground observations of solar radiation: an application in producing a five-year dataset of radiation maps on Italian territory through correction of MSG-derived data
暂无分享,去创建一个
The incident solar radiation is one of the component of the land surface energy budget and constitutes an essential input for several applications. An accurate estimation of this variable on large areas requires a dense network of ground sensors and continuous knowledge of the cloud cover, that are rarely available. A valid alternative in this respect is constituted by the remote sensing. In this work a simple algorithm is used in order to integrate the LSA-SAF (Land Surface Analysis Satellite Applications Facility) products of shortwave incident radiation obtained from MSG-SEVIRI imagery with ground radiometers observations. A statistical approach is followed in order to define a criterion for accept or reject the ground sensors observations, by modelling the mean daily error between the observations and a theoretical radiation time series and the cloud cover observations with probability distribution functions. Such distributions is used for the ground sensors selection criterion. The analysis is used to produce a dataset of corrected solar radiation maps on the whole Italian territory for a period of 5 years (2005-2009).
[1] B. Renard,et al. Use of a Gaussian copula for multivariate extreme value analysis: Some case studies in hydrology , 2007 .
[2] B. Bobée,et al. Multivariate hydrological frequency analysis using copulas , 2004 .
[3] Francesco Serinaldi,et al. Asymmetric copula in multivariate flood frequency analysis , 2006 .
[4] Sebastiano B. Serpico,et al. The OPERA project: EO-based flood risk management in Italy , 2009, 2009 IEEE International Geoscience and Remote Sensing Symposium.