Improving the operational forecasts of outdoor Universal Thermal Climate Index with post-processing

• DOI: 10.1007/s00484-024-02640-6
• Publication Date: 2024-03-05T09:02:38Z
• Authors: Danijela Kuzmanović; Jana Banko; Gregor Skok
• Keywords: klimatologija; termalni indeks; thermal comfort; Climate; Slovenia; Wind; 01 natural sciences; Machine Learning; verifikacija; Humans; Thermosensing; 0105 earth and related environmental sciences; Original Paper; post-processing; univerzalni toplotni klimatski indeks; Temperature; climatology; Humidity; info:eu-repo/classification/udc/551.58; UTCI forecasting; Linear Models; Neural Networks, Computer; Universal Thermal Climate Index; verification; Forecasting

Abstract The Universal Thermal Climate Index (UTCI) is a thermal comfort index that describes how the human body experiences ambient conditions. It has units of temperature and considers physiological aspects body. takes into account effect air temperature, humidity, wind, radiation, clothes. increasingly used in many countries as measure for outdoor conditions, its value calculated part operational meteorological forecast. At same time, forecasts UTCI tend to have relatively large error caused by forecasts. In Slovenia, there dense network stations. Crucially, at these stations, global solar radiation measurements are performed continuously, which makes estimating actual more accurate compared situation where no available. We seven years hourly resolution from 42 stations first verify forecast day and, secondly, try improve via post-processing. two machine-learning methods, linear regression, neural networks. Both methods successfully reduced daily mean about 2.6 $$^{\circ }$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mrow/> <mml:mo>∘</mml:mo> </mml:msup> </mml:math> C almost zero, while absolute decreased 5 3 3.5 regression. especially network, also substantially dependence on time day.