Abstract:The availability of high spatio-temporal resolution of urban air temperature is paramount for understanding urban heat island (UHI) and developing effective mitigation strategies, in particular for local-scale mitigations. Obtaining high spatial resolution of air temperature at city-scale is challenging as the quantity of weather stations is often limited in cities, particularly in those less developed ones. In this study, based on the existing weather station network in Guangzhou city, China, we compare eight different air temperature predictive models and select one with the best performance to interpolate city-scale air temperature. The training and validation of the models are performed using observatory meteorological data of 321 weather stations in Guangzhou. Deep learning-derived land cover information and social-economic data are encoded to be used as explanatory variables. The regression kriging combined with multiple linear regression is found to result in the best performance, with an average root mean squared error (RMSE) of 0.92 degrees C and a coefficient of determination (R2) of 0.959. Furthermore, the quantities and locations of current weather stations can be optimized by the proposed model. Guided by the k-means clustering alongside the information of geocoordinates and land cover, the number of current weather stations in Guangzhou can be reduced by 50% (i.e., 160 weather stations) while retaining the model performance. This study proposes and demonstrates an effective model for obtaining city-scale air temperature at high spatio-temporal resolution with data from sparse weather stations, which is much needed for cities which want to enhance their city-scale air temperature mapping by complementing new weather stations to their existing weather station network.

Urban microclimate prediction based on weather station data and artificial neural network

Forecasting Building Energy Consumption Using Neural Networks and Hybrid Neuro-Fuzzy System: A Comparative Study

ANN-Based Method for Urban Canopy Temperature Prediction and Building Energy Simulation with Urban Heat Island Effect in Consideration

Urban Micro-Climate Prediction Through Long Short-Term Memory Network with Long-Term Monitoring for On-Site Building Energy Estimation

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

Assessing urban micro-climates with vertical and horizontal building morphological cutting deep transfer learning neural networks

Optimization Method for Energy Saving of Rural Architectures in Hot Summer and Cold Winter Areas Based on Artificial Neural Network

Assessing built microclimate with building group frontal projection maps: A sun-path-dependent deep transfer learning neural network approach

Assessment of macroclimate and microclimate effects on outdoor thermal comfort via artificial neural network models

Hyper‐Local Temperature Prediction Using Detailed Urban Climate Informatics

Impacts of Weather Conditions on District Heat System

Predicting solar radiation in the urban area: A data-driven analysis for sustainable city planning using artificial neural networking

Thermal load prediction and operation optimization of office building with a zone-level artificial neural network and rule-based control

A neural network-based surrogate model to predict building features from heating and cooling load signatures

Study on the Impact of Building Energy Predictions Considering Weather Errors of Neighboring Weather Stations

An innovative method to predict the thermal parameters of construction assemblies for urban building energy models

Transforming and validating urban microclimate data with multi-sourced microclimate datasets for building energy modelling at urban scale

Investigating the impact of urban microclimate on building thermal performance: a case study of dense urban areas in Hong Kong

Predicting the impacts of urban development on urban thermal environment using machine learning algorithms in Nanjing, China

Impact Analysis of Urban Morphology on Residential District Heat Energy Demand and Microclimate Based on Field Measurement Data

Urban Buildings Thermal Environment Research Based On Bp Neural Network