{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,9]],"date-time":"2026-03-09T08:22:46Z","timestamp":1773044566583,"version":"3.50.1"},"reference-count":24,"publisher":"MDPI AG","issue":"8","license":[{"start":{"date-parts":[[2024,4,9]],"date-time":"2024-04-09T00:00:00Z","timestamp":1712620800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Chief Scientist of the Israeli Ministry of Agriculture","award":["20-01-0058"],"award-info":[{"award-number":["20-01-0058"]}]},{"name":"Chief Scientist of the Israeli Ministry of Agriculture","award":["18-17-0012"],"award-info":[{"award-number":["18-17-0012"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Near-surface air temperature (Ta) is a key variable in global climate studies. Global climate models such as ERA5 and CMIP6 predict various parameters at coarse spatial resolution (&gt;9 km). As a result, local phenomena such as the urban heat islands are not reflected in the model\u2019s outputs. In this study, we address this limitation by downscaling the resolution of ERA5 (9 km) and CMIP6 (27 km) Ta to 1 km, employing two different machine learning algorithms (XGBoost and Deep Learning). Our models leverage a diverse set of features, including data from satellites (land surface temperature and normalized difference vegetation index), from ERA5 and CMIP6 climate models (e.g., solar and thermal radiation, wind), and from digital elevation models to develop accurate machine learning prediction models. These models were rigorously validated against observations from 98 meteorological stations in the East Mediterranean (Israel) using a standard cross-validation technique as well as a leave-one-group-out on the station ID evaluation methodology to avoid overfitting and dependence on geographic location. We demonstrate the sensitivity of the downscaled Ta to local land cover and topography, which is missing in the climate models. Our results demonstrate impressive accuracy with the Deep Learning-based models, obtaining Root Mean Squared Error (RMSE) values of 0.98 \u00b0C (ERA5) and 1.86 \u00b0C (CMIP6) for daily Ta and 2.20 \u00b0C (ERA5) for hourly Ta. Additionally, we explore the impact of the various input features and offer an extended application for future climate predictions. Finally, we propose an enhanced evaluation framework, which addresses the problem of model overfitting. This work provides practical tools and insights for building and evaluating Ta downscaling models. The code and data are publicly shared online.<\/jats:p>","DOI":"10.3390\/rs16081314","type":"journal-article","created":{"date-parts":[[2024,4,9]],"date-time":"2024-04-09T08:55:12Z","timestamp":1712652912000},"page":"1314","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":7,"title":["Comparing ML Methods for Downscaling Near-Surface Air Temperature over the Eastern Mediterranean"],"prefix":"10.3390","volume":"16","author":[{"given":"Amit","family":"Blizer","sequence":"first","affiliation":[{"name":"Department of Geography and Environment, Bar-Ilan University, Ramat Gan 5290002, Israel"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-5158-7372","authenticated-orcid":false,"given":"Oren","family":"Glickman","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Bar-Ilan University, Ramat Gan 5290002, Israel"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7594-5277","authenticated-orcid":false,"given":"Itamar M.","family":"Lensky","sequence":"additional","affiliation":[{"name":"Department of Geography and Environment, Bar-Ilan University, Ramat Gan 5290002, Israel"}]}],"member":"1968","published-online":{"date-parts":[[2024,4,9]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"109626","DOI":"10.1016\/j.agrformet.2023.109626","article-title":"A bibliometric analysis on drought and heat indices in agriculture","volume":"341","author":"Alilla","year":"2023","journal-title":"Agric. For. Meteorol."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"1609","DOI":"10.1002\/jsfa.12204","article-title":"A Review of Smart Agriculture and Production Practices in Japanese Large-Scale Rice Farming","volume":"103","author":"Li","year":"2022","journal-title":"J. Sci. Food Agric."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Dandrifosse, S., Jago, A., Huart, J.P., Michaud, V., Planchon, V., and Rosillon, D. (Smart Agric. Technol., 2024). Automatic quality control of weather data for timely decisions in agriculture, Smart Agric. Technol., in press.","DOI":"10.1016\/j.atech.2024.100445"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1002\/joc.6570","article-title":"Estimating near-surface air temperature across Israel using a machine learning based hybrid approach","volume":"40","author":"Zhou","year":"2020","journal-title":"Int. J. Climatol."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"115358","DOI":"10.1016\/j.ecoenv.2023.115358","article-title":"Urban climate and cardiovascular health: Focused on seasonal variation of urban temperature, relative humidity, and PM2.5 air pollution","volume":"263","author":"Tsao","year":"2023","journal-title":"Ecotoxicol. Environ. Safety"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"2","DOI":"10.1002\/gdj3.102","article-title":"ERA5-HEAT: A global gridded historical dataset of human thermal comfort indices from climate reanalysis","volume":"8","author":"Barnard","year":"2021","journal-title":"Geosci. Data J."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"262","DOI":"10.1038\/s41597-022-01393-4","article-title":"NASA Global Daily Downscaled Projections, CMIP6","volume":"9","author":"Thrasher","year":"2022","journal-title":"Sci. Data"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"11333","DOI":"10.1029\/2017JD027973","article-title":"Synoptic circulation impact on the near-surface temperature difference outweighs that of the seasonal signal in the Eastern Mediterranean","volume":"123","author":"Lensky","year":"2018","journal-title":"J. Geophys. Res. Atmos."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"189","DOI":"10.1002\/joc.3971","article-title":"Satellite observations of land surface temperature patterns induced by synoptic circulation","volume":"35","author":"Lensky","year":"2015","journal-title":"Int. J. Climatol."},{"key":"ref_10","unstructured":"Hemond, H.F., and Fechner, E.J. (2022). Chemical Fate and Transport in the Environment, Academic Press."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"5637","DOI":"10.5194\/essd-14-5637-2022","article-title":"A global dataset of daily maximum and minimum near-surface air temperature at 1 km resolution over land (2003\u20132020)","volume":"14","author":"Zhang","year":"2022","journal-title":"Earth Syst. Sci. Data"},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Sebbar, B.-E., Khabba, S., Merlin, O., Simonneaux, V., El Hachimi, C., Kharrou, M.H., and Chehbouni, A. (2023). Machine-Learning-Based Downscaling of Hourly ERA5-Land Air Temperature over Mountainous Regions. Atmosphere, 14.","DOI":"10.3390\/atmos14040610"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Afshari, A., Vogel, J., and Chockalingam, G. (2023). Statistical Downscaling of SEVIRI Land Surface Temperature to WRF Near-Surface Air Temperature Using a Deep Learning Model. Remote Sens., 15.","DOI":"10.3390\/rs15184447"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Mouatadid, S., Easterbrook, S., and Erler, A.R. (2017, January 18\u201321). A Machine Learning Approach to Non-uniform Spatial Downscaling of Climate Variables. Proceedings of the IEEE International Conference on Data Mining Workshops, ICDMW, New Orleans, LA, USA.","DOI":"10.1109\/ICDMW.2017.49"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"5256","DOI":"10.1016\/j.asr.2023.02.006","article-title":"Evaluation of near-surface air temperature reanalysis datasets and downscaling with machine learning based Random Forest method for complex terrain of Turkey","volume":"71","author":"Karaman","year":"2023","journal-title":"Adv. Space Res."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"79","DOI":"10.1007\/s12665-022-10180-8","article-title":"Estimation of air temperature using the temperature\/vegetation index approach over Andhra Pradesh and Karnataka","volume":"81","author":"Arumugam","year":"2022","journal-title":"Environ. Earth Sci."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"1131","DOI":"10.1175\/2011BAMS3160.1","article-title":"Detection of Finescale Climatic Features from Satellites and Implications for Agricultural Planning","volume":"92","author":"Lensky","year":"2011","journal-title":"Bull. Amer. Meteor. Soc."},{"key":"ref_18","unstructured":"(2023, March 26). 10 & 1-Minutes Data (API). Israel Meteorological Service, Available online: https:\/\/ims.gov.il\/en\/ObservationDataAPI."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1937","DOI":"10.5194\/gmd-9-1937-2016","article-title":"Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization","volume":"9","author":"Eyring","year":"2016","journal-title":"Geosci. Model Dev."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"74","DOI":"10.1038\/s41597-021-00861-7","article-title":"Worldwide continuous gap-filled MODIS land surface temperature dataset","volume":"8","author":"Shiff","year":"2021","journal-title":"Sci. Data"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"113732","DOI":"10.1016\/j.rse.2023.113732","article-title":"Comparison of gap-filling methods for producing all-weather daily remotely sensed near-surface air temperature","volume":"296","author":"Mo","year":"2023","journal-title":"Remote Sens. Environ."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"18","DOI":"10.1016\/j.rse.2017.06.031","article-title":"Google Earth Engine: Planetary-scale geospatial analysis for everyone","volume":"202","author":"Gorelick","year":"2017","journal-title":"Remote Sens. Environ."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"193","DOI":"10.1080\/19401493.2018.1498538","article-title":"Advanced machine learning techniques for building performance simulation: A comparative analysis","volume":"12","author":"Chakraborty","year":"2018","journal-title":"J. Build. Perform. Simulation"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"733","DOI":"10.1002\/joc.3370120706","article-title":"The horizontal and vertical extension of the Persian Gulf pressure trough","volume":"12","author":"Bitan","year":"1992","journal-title":"Int. J. Climatol."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/16\/8\/1314\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T14:25:29Z","timestamp":1760106329000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/16\/8\/1314"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,4,9]]},"references-count":24,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2024,4]]}},"alternative-id":["rs16081314"],"URL":"https:\/\/doi.org\/10.3390\/rs16081314","relation":{},"ISSN":["2072-4292"],"issn-type":[{"value":"2072-4292","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,4,9]]}}}