Downscaled night air temperatures between 2030 and 2070: The case of cities with a complex- and heterogeneous-topography

Abstract

Global and regional climate models tend to underestimate temperatures over urban areas. This limitation is exacerbated in complex-topography regions, where coarse spatial resolutions of climate models fail to detect large apparent change in cities' microclimates. In this study, a night air temperature downscaling approach was proposed in cities with heterogeneous- and complex-topography. It was developed based on six ground-based national weather station datasets along the 1-km remote-sensing-based MODIS land surface temperature (LST) data, between 2000 and 2020. An ordinary least square (OLS) statistical test was applied, including the 1-km MODIS normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), evapotranspiration (ET) and the digital elevation model (DEM) datasets as independent variables. Based on the generated regression relationship and a five-year pixel-based trend analysis, the 25-km NASA NEX-GDDP datasets were downscaled into 1-km spatial resolution night air temperatures between 2030 and 2072 and under two scenarios (i.e. RCP 4.5 and 8.5), averaged from 21 CMIP5 models. Validations were carried out showing good results with an RMSE of 0.49 °C, an AME of 0.42 °C and an R-squared value of 89.5%.Further assessment was made over the largest cities in Lebanon, namely Beirut, Tripoli, Zahle and Baalbeck. Main results showed that, on average, urban night air temperatures under the intermediate RCP 4.5 scenario are expected to increase by 0.12 °C or 0.61% in 2030, 0.36 °C or 1.82% in 2050 and 0.6 °C or 3.07% in 2070. Under the worst-case climate change RCP 8.5 scenario, the average increase would be an overwhelming 7.41 °C or 38.4% by 2070. City and policy makers are invited to implement the proposed straightforward and validated downscaling approach and to use the findings for a more responsive urban planning and design to the particular environemntal and geographic conditions; this isespecially required in cities and agglomorations dominated by complex- and heterogeneous-topographic characteristics.