Long-Term variability of aerosol optical depth, dust episodes, and direct normal irradiance over Kuwait for CSP Applications

Two sources of aerosol data (from ground-based sunphotometry and long-term reanalysis), as well as irradiance data (from ground-based radiometric measurements) are investigated here. The daily, seasonal and interannual variability of aerosol optical depth is evaluated over Kuwait. Based on the MERRA-2 reanalysis, long-term aerosol trends are also established over the period 1980-2016, showing a slight increase in aerosol optical depth (AOD) since about 2000. This is conducive to a concomitant decrease in the direct normal irradiance (DNI) resource of ≈2% per decade, which can affect concentrating solar power (CSP) projects over the long-term. When compared with sunphotometer data from two AERONET stations located at a distance of only 90 km, but in somewhat differing environments, shortcomings are found in the aerosol data from the MERRA-2 reanalysis. Both bias and scatter are found in the hourly and daily AOD data, as well as occasional mismatch in the prediction of the more-or-less frequent AOD spikes caused by dust storm episodes. The use of MERRA-2 aerosol data for the prediction of clear-sky DNI with a high-performance irradiance model results in underestimation (of ≈13% on average) and substantial scatter on a 1-min basis, based on a comparison with co-located, high-quality DNI data.