The analysis of static boundary condition in solar resource assessment by satellite: the role of high-resolution Digital Terrain Model in irradiance downscaling process

The morphological characteristics of the ground have a strong influence in solar resource assessment, especially in terrain with complex orography. The use of accurate Digital Model Terrain (DTM) plays a key role in assessing the amount of solar energy that reaches the earth surface. Complex terrain and areas with local micro-climatic characteristics introduces significant intra-cell variations of irradiance compared to satellite-derived measurements affected by limited spatial resolution. This is mainly due to terrain-related effects such as elevation gradient, obstruction and shadowing by horizon (natural or antropic). Influence of terrain features has been already described by several authors (Ruiz-Arias et al, 2010, Bosch et al, 2010). These terrain-related features (elevation, terrain aspect and slope, horizon angles and sky view factor) can be considered as static boundary condition as they do not change in time. In the framework of the SolarCloud project (funded by the Italian Ministry of Economic Development) this activity has the objective to quantify the dowscaling improvement in satellite-derived irradiance at territory scale using DTM rising resolution products. It is acknowledged that surface morphology is just one of the ancillary data in addition to the satellite observations that are needed in order to get more accurate result in the solar energy assessment. For this purpose, we used satellite-derived irradiance processed by AMESIS model (Advanced Model for the Estimation of the Surface Solar Irradiance, Geraldi et al, 2012) at low (3 km) and high resolution (1 km). These data are downscaled using different DTM resolution and compared with ground based stations measurements. The analysis has been developed using R open source software exploiting SRTM3 (Shuttle Radar Topography Mission 3 Arc-Second Global), SRTM1 (respectively with cell size of ~90 m and ~30 m) and higher resolution DTM generated by the local government (with cell size of 20x20 m and 5x5 m). The horizon elevation angle and sky view factor have been computed up to 25 km distance for each grid point and azimuthal angle (at 1 degree resolution). Issues related with computational time have been addressed by implementing into the algorithm a floating window which subdivided the whole study area in subareas with increasing resolution.