An analysis of the 3D shape of cities with respect to Urban Heat Island effect

Abstract

The aim of this study was to explore how building shapes interact with urban policies and regulations. We plan to develop urban planning strategies aimed at mitigating Urban Heat Island effect as well as the impacts of, climate change to urban areas. We explore the relationship between variables like urban morphology, building heights, vegetation, green distribution, shadow and the UHI effect/ climate change (H M Abdul Fattah, 2024, p2) Research has shown that urban areas tend to have higher temperatures compared to the surrounding rural areas. This effect is known as the urban heat island (UHI) [1]. The temperature increase within urban areas is known to lead to various problems including human health issues, increase in energy consumption thereby increasing greenhouse gas emissions, as well as further contributions to climate change. To reduce these effects there is need to identify factors related to UHI and take appropriate actions to mitigate its effect in urban areas. Work done by Mr. H M Abdul Fattah on this project [8] consisted of downloading city shape file, building height dataset, calculating city wise area vs building height distribution and classifying cities into one of the four categories. Dataset about the heights of buildings globally (GHS-BUILT-H - R2023A) was obtai ned from GHSL - Global Human Settlement Layer dataset[13] .  Additionally, shapefiles, defining urban areas , were obtained from World Urban Areas, LandScan, 1:10 million (2012) [14] . Data of heights of buildings was filtered using shape files and analysed to find the distribution of building heights in the city. Based on the shapes of distribution (their skewness and dip statistic values), the shape of city was classified as unimodal right skew (Pyramid) shape, unimodal left skew (Inverse Pyramid) shape, or has no significant skew (Diamond) shape [15] [8]. In case of presence of multimodality, indicated by the dip statistic, shape was classified as "Hourglass." The "Pyramid" class signifies a concentration of shorter buildings, while "Inverse Pyramid" indicates a concentration of taller buildings. "Diamond" suggests a balanced distribution of building heights, while "Hourglass" implies varying patterns in building height distribution.

Introduction

Urban Heat Island (UHI) leads to increased vulnerability to human health issues like heat strokes, exhaustion, suicidal tendancies. It is also impacting air quality due to more amount of pollutants released in air and poor scattering of these pollutants. Water quality also gets impacted due to increase in water tempaerature affecting native aquatic life.[3]

Urban areas are densely populated with more people. Closely constructed building and skyscrapers mean a lot of waste energy is emitted and can not escape the area. Increase in temperature also causes increase in energy consumption thereby increasing greenhouse emissions for the city dwellers. [2]. According to David L. Chandler, Urban heat island effects also depend on a city’s street and building layout Some cities, such as New York and Chicago, are laid out on a precise grid, like the atoms in a crystal, while others such as Boston or London are arranged more chaotically, like the disordered atoms in a liquid or glass. The researchers found that the “crystalline” cities had a far greater buildup of heat compared to their surroundings than did the “glass-like” ones.[4] It is found that UHI is positively correlated with city area. Building materials which absorb and radiate heat back into the air gets trapped in the nearby vicinity in the area densely crowded with buildings instead of spreading out evenly[5]. Hence effect of heights of skyscrapers also needs to be studied apart from the area of city. This can help in planning for urban area expansions or in new urban area developments

According to Nyuk Hien Wong, Chun Liang Tan, Dionysia Denia Kolokotsa & Hideki Takebayashi [6], Green infrastructure acts to cool the urban environment through shade provision and evapotranspiration. Typically, greenery on the ground reduces peak surface temperature by 2–9 °C, while green roofs and green walls reduce surface temperature by ~17 °C, also providing added thermal insulation for the building envelope. However, the cooling potential varies markedly, depending on the scale of interest (city or building level), greenery extent (park shape and size), plant selection and plant placement . This can be a tool for mitigating UHI

Climate change is impacting cities and their residents in many ways, from poor air quality to flooding, biodiversity loss and extreme heat. Mackres et al. [7] with the help of a dashboard provides insight into connection between climate change and urban life which will be useful for city designing in a more sustainable and nature-positive ways to mitigate climate change

Objective of this study is to determine the impact of building heights, city area expressed as city 3D shape on urban heat island effect observed in cities along with the impact of other variables like vegetation, water surfaces in the proximities of various cities spread across in the world

Remarks

This repo has the following basic structure.

├── README.md           <- The top-level README for developers.
│
├── data            
│   ├── 01_raw          
│   ├── 02_intermediate 
│   ├── 03_processed    
│   ├── 04_models       
│   ├── 05_model_output 
│   └── 06_reporting   
│   └── 07_cities_shapefile      <- Shapefile of cities around the world
│   └── 08_cities_height         <- Building heights (raster data)
│   └── 09_continents_shapefile  <- Shapefile for continents
│   └── 10_raw_Other_variables   <- Downloaded raw data files 
│   └── 11_Output_R              <- Output files of R programs
│     └── city_rasters           <- Raster files generated using R programs
│   └── 12_Output_python         <- Output files of R programs
│
├── conf               
├── docs                <- Space for documentation. Can also included conceptualization and literature review.
│
├── references          <- Data dictionaries, manuals, reference manager (e.g. EndNote) etc.
│
├── results             
│   ├── submissions     
│
├── .gitignore          <- Avoids uploading data, credentials, outputs, system files etc.
│
├── src_py                    <- Python source code for use in this project.
│
├── src_r                     <- R source code for use in this project.

List of programs and its purpose

1. Download_img_surfacewater_NDVIcsv.ipynb python program to download surface water data from Google earth engine as 6 tiles.

2. UHI_and_Other_Variables_Raster_download.Rmd R program to reads all the data files and crops the data based on the shape file, calculates means of cities and merges NDVI, UHI, surface means into csv ‘Citywise Means [0-10].csv ’,‘Citywise Mean NDVI[0-10].csv ’, ‘Citywise Mean Surface water [0-10].csv ’

3. Download_img_surfacewater_NDVIcsv.ipynb NDVI and surface data is in the form of multiple tiles. So, a python program is written to determine the list of cities included in each tile and generate CSV containing this info. And in R program only the tile containing city is opened at a time in order to reduce memory usage.

4. Download_img_UHI_surfacewater_SolarRadiation_NDVI_coord.ipynb Python program to get city geographical cordinates and save in City_coordinates.csv

5. Merge_all_mean_csv_and_ML_model_coordinates.ipynb Python program to merge Cityshape csv (‘../data/10_raw_Other_variables/combined_summary_City_Shape.csv’) created previously with this csv containing various means Also city coordinates information was added to this csv. A single file containing all means along with coordinates is created

6. UHI_EDA_Analysis.Rmd R program used for analysis using Boxplots, Anova and Kruskal-Wallis test Also logistic regression ML model was used in a python program to explore relationship between UHI and city shape, NDVI, surface water

Data Files -

Raw Data files are in data\10_raw_Other_variables folder UHI .tif file (UHI_yearly_pixel_2018.tif) and NDVI .tif files from data\10_raw_Other_variables folder\NDVI folder are not added here due to large size. They are copied in nextcloud. Also raster files are copied to nextcloud

CSV files accessed in program -

City_NDVI_files.csv in data\10_raw_Other_variables folder, City_surface.csv in data\10_raw_Other_variables folder, combined_summary_City_Shape.csv in data\10_raw_Other_variables folder, ML_input_data_coordinates.csv in data\12_Output_python folder, UHI_NDVI_Swater_Means_data.csv in data\12_Output_python folder, City_coordinates.csv in data\12_Output_python folder, Citywise Means [0 -10].csv (11 files numbered 0 to 10) in data\11_Output_R folder

References

[1] Urban heat island, Wikipedia, Available at https://en.wikipedia.org/wiki/Urban_heat_island [2] Nidhi Singh, Saumya Singh, R.K. Mall. Urban ecology and human health: implications of urban heat island, air pollution and climate change nexus. Available at https://www.sciencedirect.com/science/article/abs/pii/B9780128207307000173#:~:text=The%20UHI%20effect%20has%20caused%20an%20increase%20in,in%20nearby%20areas%20thus%20developing%20new%20ecological%20implications.  [3] Urban Heat Island (nationalgeographic.org) Available at https://education.nationalgeographic.org/resource/urban-heat-island/  [4] David L. Chandler, Urban heat island effects depend on a city’s layout, Available at https://news.mit.edu/2018/urban-heat-island-effects-depend-city-layout-0222 [5] City module - City cluster and urban heat islands (Europe), Available at https://www.pik-potsdam.de/cigrasp-2/city-module/heat-island-cluster/index.html
[6] Nyuk Hien Wong, Chun Liang Tan, Dionysia Denia Kolokotsa & Hideki Takebayashi Available at https://www.nature.com/articles/s43017-020-00129-5  [7] Mackres, E., Pool, J. R., Shabou, S., Wong, T., Anderson, J., Gillespie, C., & Alexander, S. New Data Dashboard Helps Cities Build Urban Resilience in a Changing Climate. Available at https://www.wri.org/update/new-data-dashboard-shows-climate-change-risks-in-cities  [8] H. M. Fattah Capstone Report, Available at https://github.com/datadiversitylab/city-3D-shapes/tree/main/docs/Capstone_report_for_H_M_Abdul_Fattah.pdf  [9] Urban Heat Island dataset, Various UHI datasets available at https://developers.google.com/earth-engine/datasets/tags/uhi or Direct dataset download - https://sedac.ciesin.columbia.edu/data/set/sdei-yceo-sfc-uhi-v4/data-download [10] NOAA CDR AVHRR NDVI: Normalized Difference Vegetation Index, Version 5 dataset, available at MODIS Combined 16-Day NDVI available at https://developers.google.com/earth-engine/datasets/catalog/MODIS_MCD43A4_006_NDVI [11] MCD18A1.061 Surface Radiation Daily/3-Hour dataset, Available at https://developers.google.com/earth-engine/datasets/catalog/MODIS_061_MCD18A1
[12] JRC Global Surface Water Mapping Layers, v1.4 dataset. Available at https://developers.google.com/earth-engine/datasets/catalog/JRC_GSW1_4_YearlyHistory [13] GHSL - Global Human Settlement Layer. Available at https://human-settlement.emergency.copernicus.eu/datasets.php (GHS-BUILT-H - R2023A dataset used for building height calculation) [14] Kelso, N.V. and Patterson, T. (2012). World Urban Areas, LandScan, 1:10 million (2012). Available at https://geo.nyu.edu/catalog/stanford-yk247bg4748 [15] Elsen, PR., & Tingley, MW. (2015). Global mountain topography and the fate of montane 296 species under climate change. Nature Climate Change, 5(8), 772-776 Available at https://www.nature.com/articles/nclimate2656 [16] One-Way ANOVA Test in R - Easy Guides - Wiki - STHDA Available at https://www.sthda.com/english/wiki/one-way-anova-test-in-r [17] Kruskal-Wallis Test in R - Easy Guides - Wiki - STHDA Available at https://www.sthda.com/english/wiki/kruskal-wallis-test-in-r [18] Correlation matrix with ggally – the R Graph Gallery Available at https://r-graph-gallery.com/199-correlation-matrix-with-ggally.html

Acknowledgment

Initial project structure was created following the structure in this repo.