Abstract:The growing impact of climate change, including extreme weather events, represents a significant challenge for humanity. With most of the world's population living in urban areas, the urban heat island effect and anthropogenic heat contribute to elevated city temperatures. This increase in urban warming threatens human health and demands a deeper understanding of thermal distribution in urban environments. Collecting accessible and widespread temperature data in urban areas is essential to address this challenge. This study aims to develop a methodology for anticipating temperature distribution in urban environments, leveraging Citizen Weather Stations (CWS) as valuable crowdsourcing data sources. The ultimate goal is to create a predictive model that estimates urban temperatures based on government meteorological station forecasts, improving urban planning, regulating temperature-based routes, preventing health issues in vulnerable populations, and enhancing urban livability. The methodology is divided into three fundamental stages: data acquisition through CWS with citizen collaboration, the development and evaluation of optimal forecast models based on government weather stations (SWS) data, and its exploitation in terms of utility and applicability. This methodology encompasses data collection and filtering to ensure its usefulness and implement reliable models. The resulting tool facilitates informed decision-making and precise seasonal event planning in urban environments, effectively addressing the challenges of climate extrapolation and contributing to more effective adaptation and mitigation strategies in climate change and heatwaves. The results obtained probe the feasibility of using CWS to predict temperatures in urban environments, which has been demonstrated accurately. This is a significant achievement, as CWS has proven to be a reliable source of climate data for this context. Also, the filtering process described and applied to the case study has proven effective, discarding approximately 34.87 % of the data. This is achieved by detecting and eliminating anomalies, considering station availability, and adhering to specific quality criteria. Finally, the developed prediction model has demonstrated its ability to optimally estimate urban temperatures, utilizing climate prediction data provided by government weather stations (SWS). The model performance indicators support this claim. For the linear regression model, a Mean Squared Error (MSE) of 2.177 and an R-squared (R 2 ) of 0.960 are obtained, while for the neural network, an MSE of 1.284 and an R 2 of 0.976 are achieved.

Maximum Temperature Prediction Using Remote Sensing Data Via Convolutional Neural Network

Machine learning-assisted mapping of city-scale air temperature: Using sparse meteorological data for urban climate modeling and adaptation

A Machine Learning Approach for the Efficient Estimation of Ground-Level Air Temperature in Urban Areas

An integrated cyberGIS and machine learning framework for fine-scale prediction of Urban Heat Island using satellite remote sensing and urban sensor network data

Prediction of Urban Thermal Environment Based on Multi-Dimensional Nature and Urban Form Factors

Urban Air Pollution Forecasting: a Machine Learning Approach leveraging Satellite Observations and Meteorological Forecasts

A New Approach for Understanding Urban Microclimate by Integrating Complementary Predictors at Different Scales in Regression and Machine Learning Models

Chasing the Heat: Unraveling Urban Hyperlocal Air Temperature Mapping with Mobile Sensing and Machine Learning

Using Long Short-Term Memory (LSTM) and Internet of Things (IoT) for localized surface temperature forecasting in an urban environment

Interpolation, Satellite-Based Machine Learning, or Meteorological Simulation? A Comparison Analysis for Spatio-temporal Mapping of Mesoscale Urban Air Temperature

Modelling long-term thermal comfort conditions in urban environments using a deep convolutional encoder-decoder as a computational shortcut

High-resolution multi-scaling of outdoor human thermal comfort and its intra-urban variability based on machine learning

Hyper‐Local Temperature Prediction Using Detailed Urban Climate Informatics

Fast prediction of spatial temperature distributions in urban areas with WRF and Temporal Fusion Transformers

Spatial-Temporal Analysis of Atmospheric Environment in Urban Areas using Remote Sensing and Neural Networks

Urban land surface temperature forecasting: a data-driven approach using regression and neural network models

Empowering urban climate resilience and adaptation: Crowdsourcing weather citizen stations-enhanced temperature prediction

Synergising Machine Learning and Remote Sensing for Urban Heat Island Dynamics: A Comprehensive Modelling Approach

An improved machine learning-based model for prediction of diurnal and spatially continuous near surface air temperature

Predicting Temperature of Major Cities Using Machine Learning and Deep Learning

Insights into the Effect of Urban Morphology and Land Cover on Land Surface and Air Temperatures in the Metropolitan City of Milan (Italy) Using Satellite Imagery and In Situ Measurements