Eastvale Heat Plan: Cool Roofs, Trees or Electric Buses?
Eastvale is a city of 1.15 million people in a subtropical valley. Mean summer maximum temperature is 34.2°C, and the urban heat island intensity reaches 4.6°C at night in dense central wards, compared with 1.1°C in garden suburbs. About 38% of the city is impervious surface; tree canopy is 9% citywide but only 3% in the lowest-income eastern wards. Heat-related ambulance callouts rose from 410 in 2015 to 980 in 2024, and 71% of excess deaths in the 2023 heatwave were people over 70 living in wards with under 5% canopy. The city has a $22 million capital budget and three options: (A) 120,000 m² of cool roofs and reflective pavement in the eastern wards, costing $14M, with an estimated 1.0–1.4°C daytime reduction; (B) 18,000 street trees, costing $19M including irrigation and eight years of maintenance, with an estimated 0.6–1.8°C street-level benefit that matures in 12–15 years; (C) electric bus fleet expansion, costing $21M, cutting transport emissions by 24% but with no measurable local temperature effect by 2035. The mayor wants a visible win before the next election in 20 months, while the health department wants action for over-70s this summer.
1. Recommend an allocation of the $22 million and justify it with reference to mitigation, adaptation and political timescale.
Read the worked answer
Recommendation: $14M for cool roofs and reflective surfaces in the eastern wards, $7M for targeted shade and trees around clinics, schools and bus stops, and $1M for monitoring. Do not fund the bus expansion from this budget.
Adaptation first, because the harm is now. The health data show excess mortality concentrated in over-70s in low-canopy wards; that is a present-day adaptation problem. Cool roofs act within one hot season, are cheap per m², and reduce indoor heat, which matters most for people who cannot leave home. The estimated 1.0–1.4°C daytime reduction is modest but targeted at the most exposed wards.
Mitigation is real but slower and global. The bus fleet cuts transport emissions by 24% and is legitimate climate policy, but it will not measurably lower street temperatures by 2035. In a 20-month political horizon it is invisible, and it spends 95% of the budget on a benefit that is not local. That does not make it bad; it makes it a different budget line.
Trees are the better long-run adaptation. Option B's 0.6–1.8°C street-level benefit is plausible, but maturity in 12–15 years is too slow for this summer. A partial tree programme buys shade where people queue and walk, and it addresses the 3% canopy inequity. Full B would consume $19M and leave no monitoring or roof budget.
Equity and co-benefits. The eastern wards have 38% impervious surface and the lowest canopy, so the same spend buys more heat reduction there. Cool roofs also cut air-conditioning demand, lowering bills and emissions. Reflective pavement needs care: it can raise glare and pedestrian radiant load, so use it on low-traffic surfaces, not playgrounds.
Political timescale. The mayor's 20-month horizon aligns with the roof programme, which can be procured and visible before summer. The health department's priority is the same. The judgement is a portfolio weighted to fast, local, equitable adaptation, with mitigation deferred to a separate funding stream.
2. Explain why the 4.6°C urban heat island figure cannot be used to predict the temperature benefit of option A.
Read the worked answer
UHI intensity is a measured difference, not a property of a material. The 4.6°C is the difference between a dense central station and a garden-suburb reference at a particular time, usually at night under calm, clear conditions. It depends on the station pair, sensor height, wind, cloud, humidity, waste heat and street geometry. It is not the amount by which a roof treatment will cool the air.
Cool roofs change the surface energy balance, not directly air temperature. Increasing albedo reduces absorbed shortwave radiation and lowers surface temperature. Air temperature at 2 m responds through sensible heat flux, which depends on wind speed and mixing. On a windy day the cooled air is advected away; at night the effect may be smaller because shortwave reflection is irrelevant. Reflective pavement can also reduce surface temperature while increasing reflected radiation onto pedestrians.
Scale mismatch. A 120,000 m² roof programme is a patchwork within a 1.15 million-person city. The city-scale air temperature effect may be much smaller than the local surface effect, and the UHI figure is a city-scale difference. The 1.0–1.4°C estimate is therefore a model output with assumptions, not a deduction from 4.6°C.
Confounding. If the roofs are installed in the eastern wards, which have different building density, income and vegetation, a before-after comparison confounds treatment with weather and with other changes. The correct prediction needs a pilot with paired control sites, surface-temperature satellite data and 2 m air-temperature sensors. The 4.6°C figure can justify targeting, but it cannot quantify the benefit.
3. Design a two-year monitoring scheme to evaluate the chosen option. Explain how the data would support or overturn the decision.
Read the worked answer
Design: a difference-in-differences evaluation with paired sites. Select ten treated roofs or blocks in the eastern wards and ten control blocks matched on building height, density, income and baseline canopy in wards not yet treated. Install shielded temperature and relative-humidity loggers at 2 m in each, plus one rooftop surface sensor per pair, recording every 15 minutes for two summers. Add mobile transects at 14:00 and 02:00 on ten forecast heatwave days, using a calibrated sensor on a bicycle at a fixed route and speed.
Outcomes. Primary: 2 m air temperature difference between treated and control, daytime and nighttime separately. Secondary: surface temperature from Landsat or Sentinel on clear-sky dates, ambulance callouts for 65+ per 100,000, indoor temperature in a sample of 40 homes, and roof condition. Pre-register the heatwave definition (for example, daily maximum at least 34.2°C for two days) to avoid cherry-picking.
Analysis. Difference-in-differences: change in treated minus change in control, with 95% confidence intervals. Adjust for wind, humidity and cloud. Report nighttime and daytime separately, because the mechanisms differ.
Limitations. Two years gives two summers; a cool summer weakens statistical power. Sensors can drift, so calibrate quarterly and rotate a reference logger. Mobile transects sample street air, not the whole ward. Health data are lagged and confounded by behaviour, so treat them as supporting evidence, not proof.
Decision rule. If daytime air temperature falls by at least 0.5°C in treated sites with no nighttime penalty, scale up. If only surface temperature falls, report that honestly and consider shade as the next intervention. The scheme should be able to overturn the recommendation, not just confirm it.