Title page 1

Contents 1

Abstract 1

Introduction 2

Methodology 9

Limitations 39

Findings 40

Conclusions 41

Endnotes 41

References 41

Acknowledgments 44

About the authors 44

Table 1. Cities engaged in our user research process to understand needs for long-term adaptation to rising temperatures 9

Table 2. OpenUrban LULC category definitions and datasets 12

Table 3. Input datasets for OpenUrban 13

Table 4. OSM categories and category types used to identify features of interest for the OpenUrban dataset 14

Table 5. Accuracy assessment of OpenUrban for US cities 17

Table 6. Accuracy assessment of OpenUrban for global cities 17

Table 7. Comparison of overall accuracy of different land cover datasets 18

Table 8. Input datasets for measuring surface characteristics 19

Table 9. DynamicWorld LULC categories are recategorized to estimate NDVI₀ (built-up) and NDVIveg (vegetated) 21

Table 10. Summary of potentials for the Cape Town AOI scenarios 26

Table 11. Cool roof material potential 30

Table 12. OSM road category types used to classify road traffic volume 34

Table 13. Material options for reflective pavements based on traffic volume 34

Figure 1. Our framework for mapping scenarios and estimating potential for heat-resilient infrastructure, using street trees as an example 3

Figure 2. Cool Cities Lab workflow 4

Figure 3. Heat-resilient infrastructure scenario framework 8

Figure 4. OpenUrban data for Portland, Oregon (United States) 10

Figure 5. OpenUrban data for Istanbul 11

Figure 6. Roof slope classification (United States only) 15

Figure 7. Tree (left) and cool roof (right) opportunity in Cape Town, South Africa 23

Figure 8. Area of interest in Cape Town, South Africa 25

Figure 9. Street tree planting program to local best practice levels on all nonhighway roads 29

Figure 10. Park tree planting program to local best practice levels in all public open spaces 29

Figure 11. Map for program scenario cool roofs on large buildings 33

Figure 12. Map for program scenario cool roofs on all buildings 33

Figure 13. Map for program scenario reflective pavements on all low-traffic roads 36

Figure 14. Map for program scenario shade cover increase to ensure minimum target level 38

Figure 15. Combination of street tree and cool roof program scenarios 39