Investigating Actual and Future Trends of Thermal Characteristics with Satellite Images and Artificial Neural Networks Approach

Rapid and unplanned urban expansion, coupled with detrimental changes in Land Use Land Cover (LULC), lead to a degradation of the urban thermal environment and give rise to various unfavourable ecological consequences. Alterations in LULC and thermal properties carry significant implications for the economy, climate trends, and the sustainability of the environment. This research focuses on the Republic of Malta, analyzing alterations in LULC and the Urban Thermal Field Variance Index (UTFVI) spanning from 1993 to 2023, also predicting their distributions for the year 2033. To accomplish this, the analysis employs imagery from Landsat 4-5 Thematic Mapper (TM) and Landsat 8-9 Operational Land Imager (OLI). The classification of LULC is carried out using a supervised satellite image classification system, while predictions are conducted using the Cellular Automata-Artificial Neural Network (CA-ANN) algorithm. The calculation of LST employs the radiative transfer technique, with the same CA-ANN algorithm employed to forecast UTFVI for the year 2033. In order to explore the temporal correlations between LULC alterations and UTFVI, a cross-tabulation technique is employed. The findings of the study suggest that between 1993 and 2033, there will be a 9.4 % increase in Vegetation cover, predominantly at the expense of Bare-land. The Strongest UTFVI zone is anticipated to contract from 2023 to 2033, while None UTFVI class demonstrates a significant increase beyond 2.85 km(2). The predictions underscore the pressing necessity for proactive measures to conserve and safeguard the burgeoning Vegetation cover, thus upholding ecological equilibrium, mitigating the urban heat island effect, and protecting Malta's biodiversity.