Street-level greenery mapping and urban heat modelling using deep learning and street view imagery

Abstract

In an urbanizing world, challenged by climate change and rising temperatures, street vegetation can help mitigate urban temperatures. However, scaling the understanding of street vegetation’s cooling effects from local observations to the city level introduces challenges. While several studies have assessed the influence of street greenery on land surface temperature (LST) at the city scale using street view (SV) data, few have distinguished between different types of greenery. Additionally, the combined investigation of spatial non-stationarity and non-linear trends remains underexplored.
This study fine-tuned a semantic segmentation model for the city of Zurich using SV imagery. The model extracted urban features - trees, shrubs, grass, buildings, and sky - by calculating view factors (VFs). Street orientation and water presence were quantified using ancillary data. These urban characteristics served as inputs for a Geographically Weighted Random Forest (GWRF) model, complemented by SHAP analysis, to evaluate their impact on LST. Moreover, different approaches to quantifying the extracted features via VFs were assessed.
The fine-tuned deep learning model effectively captured different urban features from the SV data, achieving high Intersection over Union (IoU) scores across the extracted classes: tree (IoU = 0.91), shrub (IoU = 0.76), grass (IoU = 0.83), building (IoU = 0.92), and sky (IoU = 0.99). The GWRF model proved effective at predicting LST, with good evaluation metrics for both training (R² = 0.89, MAPE = 2.10%) and test data (R² = 0.87, MAPE = 2.26%). SHAP analysis revealed that buildings had the strongest positive effect on LST at moderate to high VFs, with a clear non-linear trend. Conversely, trees had the strongest cooling effect under similar VF ranges. Shrubs had a moderate, yet more spatially distributed cooling effect, suggesting their potential in areas where space for large canopy trees is limited.
The use of a modified, fine-tuned network proved effective in improving classification results of the street view images. In future research, this approach can be extended with additional classes to investigate the influence of specific urban characteristics. The applied method, integrating GWRF and SHAP analysis, showed potential for studying the relationship with LST while considering spatial non-stationarity and non-linear effects. By applying this method to other case studies, its wider applicability in analysing the dynamics of LST can be evaluated. Such studies could reveal spatial trends and support urban planners and policymakers in effectively mitigating LST in urban environments.

Popular Abstract

An increasing number of people are living in urban areas. At the same time, climate change and rising temperatures are posing challenges to the increasingly urbanised world, necessitating measures to mitigate these negative effects. Moreover, dense streets are particularly susceptible to high temperatures because the high density of buildings traps heat, as well as residual heat from cars and air conditioning systems. Vegetation on streets can mitigate elevated temperatures through mechanisms such as shading, reflecting solar radiation, and evapotranspiration. However, to inform urban planners and policymakers effectively, it is necessary to understand the precise cooling effects of vegetation on a city-wide scale. It is also necessary to understand not only the effect of trees, but also that of other vegetation types, such as shrubs and grasses, as well as other urban characteristics.
This study focused on investigating the impact of trees, shrubs, grasses, buildings, sky exposure, street orientation, and proximity to water on land surface temperature (LST) in Zurich. Street View (SV) data was used to quantify the effects of the first five factors by extracting these features from the collected images using a custom-trained artificial intelligence (AI) model. SV images were collected at 50 m intervals across the entire city of Zurich. Street orientation and proximity to bodies of water were calculated using street network and water body data. The relationship between these seven urban features and LST was modelled using a geospatial AI approach.
The study found that the custom-trained AI model effectively captured various urban features from the SV data. Additionally, the model used to explore the relationship between urban features and LST performed well. The study revealed that buildings and trees had the most pronounced effect on temperatures, with large trees leading to clear reductions and densely built streets leading to increases. Furthermore, the cooling effect of shrubs was less pronounced, but more spatially distributed. This could indicate that shrubs could help to mitigate temperatures where there is insufficient space for large tree canopies. However, it should be noted that this study investigated the relationship between LST and did not quantify the thermal comfort resulting from effects such as the shading provided by trees. Grasses were only associated with cooling effects in the case of sufficiently large grass patches, while street orientation had a very limited influence. Proximity to water was found to have a very noisy relationship with LST, with consistent cooling effects only observed for close proximity to large bodies of water. Finally, it should be noted that one must always exercise caution when interpreting model results based on correlation techniques. The results show how the model works, not how the real world works.

• author: Lambrecht, Toon
• supervisor: Rachid Oucheikh; Pengxiang Zhao
• organization: Dept of Physical Geography and Ecosystem Science
• course: NGEM01 20251
• year: 2025
• type: H2 - Master's Degree (Two Years)
• subject: Earth and Environmental Sciences
• keywords: Physical Geography and Ecosystem Analysis, Google Street View, Semantic Segmentation, Deep Learning, Street Greenery, Urban Heat Modelling, Geographically Weighted Random Forest, Zurich.
• publication/series: Student thesis series INES
• report number: 726
• language: English
• id: 9205613
• date added to LUP: 2025-06-25 12:52:13
• date last changed: 2025-06-25 12:52:13