Statistical Downscaling of Global Climate Projections along the Egyptian Mediterranean coast

The climatic parameters (surface air temperature, surface relative humidity, surface wind regime, and mean sea level pressure) are important in addressing adaptation/mitigation to climatic changes. In particular, the recent and future of these climatic parameters along the Egyptian Mediterranean Coast (EMC) were analyzed based on hourly real observed data (2007–2020) and hourly reanalysis (ERA5) database (1979–2020) together with daily GFDL (global climate model) mini-ensemble mean (2006–2100). Recent climatic studies in the study area have not given enough attention to the downscaling approach, underscoring the need to set up a statistical downscaling technique to better understand the forces that govern climatic change. Here, we analyze the current climatic and future scenarios for the parameters studied in a three-step process. The first step is to study the current weather variabilities in the short term (14 years) using the real observed data. The second step is to describe the long-term (42 years) current weather variabilities over the studied stations using a reanalysis ERA5 database after bias removal by comparing with the observations. The third step is to statistically downscale the GFDL mini-ensemble, which means describing the future projection along the study area up to 2100. The statistical downscale technique is built on the developed bias correction statistical model by matching cumulative distribution functions (CDF) of the mini-ensemble mean and observations during the overlapped period (2007–2020). The results show that ERA5 describes the efficiency of the weather characteristics of the five studied stations. This data, along with the EMC 2006–2020, displays a significant positive trend for surface air temperature and significant negative trends for surface wind speed, relative humidity, and sea level pressure. The GFDL mini-ensemble mean projection, up to 2100, has a significant bias with the studied weather parameters. This is partly due to the GFDL coarse resolution (2° × 2.5°). After removing the bias, the statistically downscaled simulations from the GFDL mini ensemble mean show that the study area’s climate will experience significant change, especially surface air temperature and relative humidity with a great range of uncertainties according to the scenario used and regional variations. Our results are the initial step in enhancing the understanding and development of statistical downscaling techniques to project future climate scenarios over EMC.