Development of a very high resolution solar radiation cadaster for estimating solar energy potential across the entire federal state of Salzburg, Austria

Abstract

<p>The effect of local shadows cast by e.g. infrastructure or vegetation is of paramount importance for accurately estimating the local solar energy potential. Given the high spatial and temporal resolution that is required to provide accurate estimates of solar radiation components at very fine scales, deriving a solar cadaster across large areas is a computationally demanding undertaking. Here we present the development of a solar cadaster for the Austrian federal state of Salzburg, covering an area of more than 7.150 km² at a spatial resolution of 0.5 meters and a temporal resolution of 1 hour.</p><p>The dataset is based on the radiation model STRAHLGRID, which calculates near-surface direct and diffuse solar radiation on horizontal, real and arbitrarily inclined surfaces in the spectral range 0.3-3 μm, their component sum (global radiation), and sunshine duration. Results are provided at a spatial resolution of 100 m and a temporal resolution of 15 minutes across the national territory of Austria in near real-time. The model takes atmospheric turbidity, cloudiness, terrain shading, multiple / terrain reflections and ground albedo feedbacks into account. It integrates temporal changes of atmospheric turbidity by including precipitable water (water vapor transmittance) and visibility fields (aerosol transmittance) as obtained from the nowcasting model INCA as well as a cloud raster based on measured sunshine fraction and satellite data. The underlying digital elevation model (DEM) used as input in STRAHLGRID and INCA has a spatial resolution of 100 m.</p><p>The downscaling procedure used for providing a consistent solar radiation cadaster is based on combining the 100 m radiation data with a very high resolution digital surface model (VHR-DSM) obtained from airborne laser scanning. First, a test reference year is computed by aggregating the 15 min data to one hour and averaging across all years between 2006 and 2021 on an hourly basis. Second, diffuse and direct radiation on the horizontal surface are upsampled using bilinear interpolation. Third, radiation datasets are modified based on the VHR-DSM. Diffuse radiation is corrected using an updated version of the sky view factor based on the original DEM and the VHR-DSM. Direct radiation is corrected using shade maps from the VHR-DSM derived from the horizon angle and the solar position, computed using azimuth angle steps of 1 degree and time steps of 4 minutes. In addition to applying the shadow mask, direct radiation on the real surface is obtained using a correction term for the tilted surface based on slope, aspect and solar position. Finally, global radiation is computed as the sum of diffuse and direct radiation, which is of high relevance for spatial energy planning and solar energy applications at diverse scales.</p>