Modeling soil temperature based on Gaussian process regression in a semi-arid-climate, case study Ghardaia, Algeria

The renewable energy is the best energy potential to exploit, because they are economic, not pollutant and permanent. As a kind of renewable energy, geothermic which becomes more and more widely used in this field. In a geological setting regional on the effectiveness of the process solar thermal, offering a greater supply of geothermal energy, the study of Ghardaia’s case are based on data of soil temperature and especially using local meteorological data, Accurate estimates of mean daily soil temperature (MDST) are needed. In this study, we will use the capability of Gaussian process regression (GPR) for modeling MDST using 3 years of measurement (2005–2008), in a semi-arid climate. It was found that GPR-model based on mean air temperature as input, give accurate results in term of mean absolute bias error, root mean square error, relative square error, and correlation coefficient. The obtained values of these indicators are 0.0021, 0.5036, 0.0029 and 100 %, respectively, which shows that GPR is highly qualified for MDST estimation in semi-arid climate.