<!-- ARTICLE_BODY_START --> A study published on May 6, 2026 estimated the air-temperature cooling associated with tree cover across 8,919 large urban areas. Its population-weighted summer mean was 0.15 degrees Celsius, with substantial variation among places. That number is an aggregate estimate, not the temperature difference a person should expect when stepping beneath a particular tree. The paper also found unequal distribution of cooling benefits. Its abstract offers a useful entry into a persistent confusion: a citywide air-temperature effect and immediate relief from direct sunlight describe different experiences at different scales. McDonald and colleagues, primary research abstract
A person waiting beside an exposed road may feel an obvious improvement after moving a few metres into shade while a properly shielded air thermometer records little change. A satellite may show a striking difference between dark pavement and vegetation without revealing the air temperature inside a nearby apartment. A reflective surface can become cooler while sending more sunlight toward someone standing above it. None of these observations is inherently contradictory.
The problem is the use of temperature as though it were one interchangeable quantity. Surface temperature, air temperature, mean radiant temperature and heat-stress indices are related, but they are not synonyms. A map can be accurate about the quantity it measures and misleading when used to answer a different question.
Urban heat adaptation depends on keeping those distinctions visible. A project intended to reduce a pedestrian's exposure should be evaluated where pedestrians spend time. A project intended to reduce building heat gain needs building-relevant evidence. A citywide average can help assess broad patterns, but it cannot establish whether a particular bus stop is usable in the afternoon.
This article explains the measurement problem and how to evaluate interventions without demanding that one number represent every benefit. The research findings are dated to their sources, checked on October 31, 2026. The neighbourhood, routes and calculations used for illustration are fictional. This is not individual medical advice or a substitute for local heat warnings and qualified workplace safety assessment.
What the satellite can see
NASA's explanation of land-surface temperature makes the central distinction explicit: the satellite observes the surfaces visible from above, which may include roofs, vegetation or ground, rather than measuring the air temperature reported in a weather forecast. A tree canopy can therefore occupy the observed surface while the space beneath it remains hidden from that particular view. NASA's land-surface temperature explanation
Imagine a fictional neighbourhood, Juniper Ward, with a warehouse, a small park and a row of apartments. A satellite image assigns temperatures to areas containing those features. The warehouse roof may be hot at the overpass time. That is relevant to the roof's energy balance and possibly to the building beneath it. It is not a direct measurement of the temperature at a resident's breathing height inside an apartment across the street.
The size of the measurement footprint matters as well. One pixel can combine several materials, each contributing to the observed signal. Enlarging the image on a screen does not create new measurements within that pixel. A sharply drawn boundary in a finished map may reflect processing or a separate land-cover dataset rather than thermal detail directly resolved by the instrument.
This does not make satellite data unsuitable for urban analysis. Their broad spatial coverage can reveal patterns that a handful of ground instruments would miss. The limitation is a reason to name the measured quantity and resolution, not to discard the map. The strongest use combines the large-scale view with observations designed for the question at ground level.
For Juniper Ward, a useful caption would identify the satellite product, acquisition time, units and whether the display shows actual surface temperature or a difference from a reference. It would avoid labelling the image as the temperature residents felt. The information becomes more valuable when its boundary is clear.
A processed temperature is still a measurement with assumptions
A thermal sensor records radiation, and a surface-temperature product requires additional information and processing to interpret that signal. USGS describes Landsat Collection 2 surface-temperature generation using thermal data, emissivity information and atmospheric inputs. Its documentation also identifies caveats involving missing emissivity data, small-target blockiness, vegetation adjustment and clouds or cloud shadows. These limitations belong to that specific product and version; they should not be silently transferred to every thermal dataset. USGS product documentation
For a reader, the practical lesson is that a colourful map should have a data-quality account. Blank areas might represent missing or unsuitable observations rather than unusually cool conditions. A suspicious sharp patch might be a real land-cover contrast or a processing issue. The analyst should check quality information before translating a visual feature into a claim about neighbourhood conditions.
Comparisons between dates require care. If one image was acquired under different weather, season or surface moisture conditions, a change in colour does not automatically measure the effect of a policy. A reflective roof installed between the dates could matter, but so could several other changes. The evaluation needs a comparison that separates the intervention from the changing background.
The legend is another source of avoidable confusion. Two maps can use the same red colour for different numerical ranges. A city can appear dramatically hotter simply because the display stretches a narrower interval across the palette. Conversely, a common scale may hide a locally meaningful difference. The solution is to show the scale and explain why it was chosen.
An honest map is not one without uncertainty. It is one that lets the reader identify what the data support. Juniper Ward's map might be excellent for locating sun-heated surfaces that deserve closer investigation. It would still need additional evidence before ranking residents' personal exposure or claiming a measured health improvement.
Air temperature is measured in a particular place
A conventional air-temperature measurement aims to describe the air rather than the heating of the thermometer by direct sunlight. The instrument's shielding, ventilation and placement therefore affect the result. An exposed sensor attached to a sunlit wall can respond to its own radiative environment as well as to the surrounding air.
Suppose Juniper Ward installs inexpensive sensors on ten street signs. If one sensor receives direct afternoon sun while another sits in a ventilated shield, their difference may partly reflect the installation. Mapping the readings with elaborate software will not remove that basic comparability problem. The measurement setup must be documented before the numbers can be interpreted confidently.
Height matters because conditions near a surface can differ from those higher above it. So do nearby walls, exhausts, vegetation and airflow. A sensor beside an air-conditioning outlet measures a real local environment, but the result should not automatically represent the entire block. The question is whether the location matches the intended claim.
A useful network can include both standardised reference measurements and deliberately local measurements. The reference helps compare conditions across the city. The local measurement can reveal what happens at an exposed queue, a sheltered courtyard or a worksite. Confusion arises when those purposes are mixed without explanation, not when different instruments are used for different questions.
For public reporting, each sensor should have an identifiable location category, measurement height and basic quality history. Exact personal addresses need not be disclosed when privacy is involved. The report should state whether the readings were screened for instrument problems and whether gaps remain. A highly detailed map of poorly characterised measurements can be less trustworthy than a simpler map whose limits are known.
Radiation explains why shade can feel different immediately
Air is only one route by which a person exchanges heat with the environment. Sunlight and radiation from surrounding surfaces also matter. Mean radiant temperature is a way of representing the surrounding radiative environment as an equivalent temperature under a defined formulation. It is not the air temperature, the surface temperature of one object or the body's internal temperature.
A person in Juniper Ward can receive direct sunlight, reflected sunlight and thermal radiation from several directions. A wall, pavement, sky and tree canopy occupy different parts of the person's surroundings. Moving into shade changes that arrangement, sometimes without a large immediate change in the air. The experience is therefore consistent with a substantial radiative benefit and a modest air-temperature difference.
An intuitive comparison is standing near a warm oven in a room whose air thermometer has barely changed. The analogy illustrates an additional heat-exchange pathway; an outdoor street has a more complicated combination of sources and directions. It also shows why a single surface reading cannot describe the whole radiative field around a person.
Measurements intended to estimate mean radiant temperature need to account for how radiation arrives from different directions or use an appropriate instrument and conversion. A black-globe thermometer responds to both radiation and convective exchange, so its raw reading should not casually be relabelled as mean radiant temperature. Wind and the instrument's physical properties can matter to interpretation.
For urban design, the important question is which surfaces and directions are changed by the intervention. A canopy that blocks direct sunlight can help at one time of day while leaving low-angle sunlight exposed at another. A nearby reflective wall may alter the benefit. Shade performance is a spatial and temporal relationship, not merely the presence of an overhead object on a construction plan.
A cooler pavement can create a warmer radiative setting
A 2024 study of reflective pavement in Phoenix combined field measurements with modelling. Its primary abstract reported pavement surface temperatures 10–13 degrees Celsius lower in the afternoon, while mean radiant temperature was 5–7 degrees higher around midday. The time windows and quantities differ, so the numbers should not be subtracted to calculate a single net benefit. The result illustrates why a surface-temperature reduction alone cannot establish improved pedestrian thermal conditions. It is a finding from a particular study, not a verdict on every reflective pavement. Elmagri, Kamel and Ozer, primary abstract
The energy accounting is straightforward at a conceptual level. If a surface reflects more incoming sunlight, it absorbs less of that incoming energy. The reflected energy goes somewhere. Some may reach the sky, some a building and some a person. The effect on people depends on geometry, exposure and the other heat-exchange processes operating at the time.
For Juniper Ward, a proposal to resurface a plaza should therefore state its objective. Lowering the temperature of the pavement might reduce contact-related concerns or change heat storage. Improving comfort for people standing in direct sun is a different target. The same project could help one objective while leaving another unresolved or introducing a trade-off.
A useful evaluation would measure more than the most favourable quantity. It would compare surface, air and relevant radiative conditions at representative pedestrian locations and times. It would also examine whether the result changes under shade, after sunset or as the material ages. These are questions for a site-specific assessment, not reasons to assume that the Phoenix result must recur unchanged elsewhere.
The public claim should name the result precisely. A cooler surface is an achievement if cooler surfaces are needed. It becomes misleading when promoted as a matching reduction in the temperature people breathe or the heat stress they experience without measuring those outcomes.
Shade has a schedule
The shadow cast by a structure moves as the Sun's position changes. A shelter can shade a bench in the morning and leave it exposed during the afternoon period when more passengers use it. A plan view showing a roof above a waiting area does not establish how much usable shade reaches the waiting area at the relevant time.
A 2019 primary conference abstract from the researchers behind the 50 Grades of Shade project describes hourly mobile measurements in Tempe during three summer days in 2018. They compared multiple shade types while recording air conditions, directional radiation and surface temperature. The abstract reports preliminary findings, so it should not be treated as the complete later journal study. Its useful methodological point is the explicit comparison across shade types and hours. Schneider and colleagues, EGU research abstract
At fictional Juniper Ward's bus stop, the design team could begin by observing where people actually wait. They might stand near the sign to see approaching buses, move away from a smoking area or leave the bench because it is crowded. A canopy that shades an unused patch of pavement would look successful in an overhead drawing while delivering less benefit than intended.
A shade assessment should therefore identify the occupied area, times of use and seasonal conditions. It can compare the proposed shadow with those patterns. It should also consider whether the structure blocks airflow, creates visibility problems or becomes difficult to maintain. The objective is usable relief within a functioning public space, not maximum shadow area divorced from use.
Trees add another timescale because their size and canopy condition change. A newly planted tree is not equivalent to the mature canopy shown in a rendering. An honest project report distinguishes immediate protection, expected future shade and the care needed for that future benefit to occur. The difference is important for people using the site during the years between planting and maturity.
Humidity changes the meaning of a hot day
Evaporation is one route by which the body can lose heat. The surrounding air's moisture conditions affect the potential for that process. This is why air temperature alone cannot describe all aspects of environmental heat stress. The same measured air temperature can accompany different evaporative conditions.
The wet-bulb concept is useful here, but it is often misused in public discussion. A wet-bulb temperature reflects the behaviour of a wetted sensor under specified conditions. Natural wet-bulb measurements used in some heat-stress instruments are not simply interchangeable with every calculated psychrometric wet-bulb value. Measurement setup and assumptions matter.
A number called wet-bulb temperature is also not the same as wet-bulb globe temperature. The latter is an index that combines measurements representing several environmental influences. Confusing the terms can lead a reader to compare a threshold or finding from one quantity with a reading of another. Units in degrees do not make the quantities identical.
For Juniper Ward, this means a heat dashboard needs a clear data dictionary. If a panel displays air temperature, relative humidity, heat index and WBGT, each should have its own label and explanation. A generic feels-like heading may conceal differences in the intended use of the products. A user should not need to infer the formula from the colour scale.
The personal significance of environmental heat also depends on activity, clothing, health and other factors. No single number in this article establishes a universal safe exposure time. A city can improve how it measures and communicates environmental conditions while still directing people to appropriate public-health guidance and employers to qualified occupational assessments. Better measurement narrows uncertainty; it does not remove human variation.
Heat index and WBGT are not rival thermometers
The US National Weather Service explains that heat index combines air temperature and relative humidity under shaded conditions, while its WBGT information incorporates additional influences such as wind and solar exposure. The page presents an outdoor WBGT formulation using natural wet-bulb, globe and dry-bulb temperatures. These products serve different purposes and should not be compared as though one were a corrected version of the same thermometer reading. NWS comparison of WBGT and heat index
An original arithmetic illustration helps make the distinction visible. Using the outdoor expression 0.7 times natural wet-bulb temperature, plus 0.2 times globe temperature, plus 0.1 times dry-bulb temperature, hypothetical component values of 25, 45 and 35 degrees Celsius give 30 degrees Celsius WBGT. That calculated value is an index constructed from those components. It is not a claim that the air cooled from 35 to 30 degrees.
If the hypothetical globe value fell to 35 while the other two values stayed fixed, the index would become 28. The example isolates the arithmetic contribution of one component; in a real environment, several components may change together. It is not an instruction for judging safe work, exercise or individual health.
The example also shows why index differences should be reported in their own terms. A two-degree WBGT change is not automatically equivalent to a two-degree air-temperature change, and neither is automatically a two-degree change in body temperature. A project can report improvement in a defined index while remaining clear about the assumptions and measurements behind it.
For a city reader, the useful question is which product fits the decision. A regional warning product provides broad context. A local site assessment can address conditions that the regional product does not resolve. The two are complementary when their scope is understood. Choosing the lowest available number from several indices would defeat the purpose of having more information.
Work changes the heat balance
OSHA's heat-hazard guidance identifies environmental conditions, workload, clothing and personal factors as relevant to occupational heat stress. It recommends on-site assessment using WBGT and explains why a distant weather report may not represent a particular work location. These are occupational guidance statements, not a complete legal determination about any workplace, and the article does not reproduce exposure limits or prescribe a work-rest schedule. OSHA's heat-hazard recognition guidance
The broader measurement lesson is that a worker and a seated observer can encounter the same surroundings while producing different amounts of internal heat. A street-level environmental map describes only part of that difference. Protective clothing can further change heat exchange. A map that assigns every person the same risk based solely on location would ignore those distinctions.
At Juniper Ward, a shaded delivery area could be a useful improvement while still requiring a workplace assessment. A person unloading heavy goods has different demands from someone briefly walking through. The design team should not present an environmental improvement as proof that all activities at the site are safe under all conditions.
This distinction also matters in evaluations. If a study asks participants how hot they feel, the answer can depend on what they were doing before the measurement, how long they stayed and what they wore. Those details are part of the study context. A comparison between groups performing different tasks cannot automatically isolate the effect of the street design.
Human experience belongs in heat research, but it needs careful interpretation alongside physical measurements. A comfortable rating does not certify the absence of physiological strain, and a high surface reading does not reveal a person's internal temperature. The aim is to combine appropriate evidence for the question, rather than ask one measure to stand in for another because it is easier to collect.
A route is an exposure pattern
People move through cities. Their exposure is not necessarily the temperature at their home address or the average of their neighbourhood. A short exposed wait, a shaded walk and time in a cooled building create a sequence. Mapping only one location can miss the part of the day that matters most to the decision.
Imagine two fictional routes from Juniper Ward's apartments to a clinic. One is shorter but includes a long unshaded wait at a crossing. The other is longer but has continuous shade and places to rest. An app that optimises distance alone will prefer the first. An app that colours streets by one afternoon surface image may also choose poorly if the route is used at a different time.
A useful route assessment would connect walking time, waiting time, shade and relevant environmental measurements. It would identify barriers that prevent people from using the apparently better route. A longer route may be impractical for someone with limited mobility or may cross an unsafe road. Thermal benefit cannot be evaluated independently of access and usability.
Time weighting can help describe the sequence, but it has limits. Averaging a heat index across a journey does not automatically create a validated physiological dose. Peaks, recovery opportunities and individual circumstances may matter in ways that a simple average hides. The calculation should be labelled as a descriptive summary unless a specific validated interpretation is available.
For planning, even a careful descriptive account can be valuable. It may reveal that shading a waiting point serves more people than adding isolated trees along a rarely used segment. It may show that a nominally short trip includes substantial stationary exposure. The aim is to improve the real sequence people experience, not merely the appearance of a map divided into administratively convenient blocks.
Averages can erase the people a project is meant to help
A citywide average gives each included observation or area a particular weight. Population-weighted, land-area-weighted and pedestrian-time-weighted summaries answer different questions. A park with few users can occupy a large area; a crowded transit stop can occupy very little. The chosen weighting determines how those places contribute to the headline.
Consider an invented evaluation with two locations. One serves ten people for ten minutes each, and the other serves one hundred people for twenty minutes each. The first represents one hundred person-minutes of use; the second represents two thousand. Giving the locations equal weight can be appropriate for comparing site performance, but it does not describe how many people or how much time the improvement reaches.
This does not mean the busier location always deserves priority. The less-used site might serve people with fewer alternatives or a particularly important function. The point is that the allocation contains a value judgement as well as a measurement. A transparent plan identifies the criteria rather than allowing an averaging method to make the decision invisibly.
Juniper Ward's evaluation could report both physical performance and use. It might describe the change in a measured quantity at each site, the hours of effective shade, observed occupancy and access constraints. No single combined score is necessary if the separate results are understandable. Combining everything into one index can make trade-offs harder to inspect.
Equity also requires attention within neighbourhoods. A block-level average can conceal differences between homes, workplaces and outdoor waiting areas. Residents who cannot shift their schedules may experience conditions that an average daytime visitor avoids. Good measurement does not decide every policy priority, but it can make those differences visible enough for the policy discussion to be honest.
Night-time conditions deserve their own measurements
A daytime image cannot describe the entire thermal history of a place. Surfaces that absorb energy can later release it, while buildings and airflow create different conditions after sunset. The absence of direct sunlight changes the balance among heat-exchange processes. A design that looks favourable at one hour may perform differently at another.
For Juniper Ward, an intervention evaluation should specify whether its objective concerns daytime use, evening activity, overnight conditions or several periods. A shaded playground and a bedroom that remains warm overnight involve different questions. Measuring only the playground surface at noon cannot establish what happened inside homes during the night.
The temporal pattern matters even when only outdoor conditions are considered. An exposed plaza may cool differently from a narrow courtyard. A tree canopy changes both daytime solar exposure and the view of the sky. The net result depends on the setting, so the article does not assign a universal nighttime advantage to every shaded or unshaded form.
A practical monitoring schedule can sample the periods relevant to actual use rather than distribute effort evenly for convenience. If a location is crowded in the early evening, those hours deserve attention. If the concern is a sequence of hot days, the evaluation should not rely solely on one isolated afternoon with different preceding conditions.
The report should also distinguish a measured nighttime effect from an expected one. A project may plausibly reduce stored heat, but plausibility is not the same as observing the resulting conditions. Describing the mechanism can justify what to measure next. It should not replace the measurement when the public claim is that the benefit has already been delivered.
Indoors is not a smaller version of outdoors
A resident's indoor environment depends on the building as well as on outdoor weather. Roofs, walls, windows, ventilation, internal heat sources and cooling equipment shape the conditions inside. A satellite view of the roof and a street-level air sensor provide relevant context, but neither directly measures a room occupied by a person.
Suppose Juniper Ward's coolest-looking block contains apartments with limited ventilation and no effective cooling. Another block has warmer outdoor surfaces but buildings with better protection and reliable cooling. A surface-only ranking might reverse the ordering of indoor exposure. This is a hypothetical example of why the question must be specified, not a claim about any actual neighbourhood.
Indoor measurement introduces practical and ethical issues. Residents need to understand what is being recorded and how the information will be used. Temperature records can reveal occupancy patterns or other details when combined with additional data. A study should minimise unnecessary personal information and avoid making participation a condition for receiving assistance.
Comparisons should also account for how spaces are used. A vacant room with closed blinds is not necessarily comparable with a busy room containing appliances. A measurement taken while cooling equipment is running answers a different question from one taken during an outage. These are not reasons to abandon indoor studies; they are reasons to record the context.
For policy, a useful heat strategy may combine public-space improvements with building-focused measures and access to appropriate services. The evidence for each component should match its intended outcome. A city should not claim that planting street trees has solved indoor exposure without examining the buildings, just as an efficient building does not automatically provide relief to a person waiting outside it.
Measuring an intervention requires a believable comparison
A cooler day after construction does not by itself establish that construction caused the cooling. Weather changes, seasonal differences and nearby alterations can affect the result. An evaluation needs a comparison that helps separate the intervention from those background changes.
An original arithmetic example uses air temperature at two fictional sites. Before a project, the treatment site is 34 degrees and a comparison site is 33. Afterward, the treatment site is 32 and the comparison site is 32. The treatment site cooled by two degrees, but the comparison cooled by one. A simple difference-in-differences calculation attributes a one-degree relative change to the treatment contrast, subject to its assumptions. It is not proof of causality by arithmetic alone.
The credibility depends on whether the comparison site would have followed a similar trend without the intervention and whether other relevant changes occurred. If one site received new irrigation or its sensor moved, the comparison may be compromised. The report should identify these possibilities rather than present the subtraction as a substitute for study design.
Repeated measurements across relevant conditions can make the result more informative. A project may have a strong effect under clear skies and a smaller effect under clouds. That variation can be useful if it is reported as part of the result. Selecting only the most favourable afternoon creates a performance claim that users cannot reliably apply.
A good evaluation also names its primary outcome before inspecting the results. If the objective is pedestrian radiative relief, a favourable pavement reading should not replace an inconclusive pedestrian measurement after the fact. Exploratory findings can be valuable, but they should be labelled as such and tested appropriately rather than quietly promoted into the original goal.
Mobile surveys need a clock as well as a map
A mobile instrument can cover many streets and reveal spatial contrasts. It also observes different places at different times. If the city warms while the survey moves, part of an apparent geographical pattern may be a temporal trend. The order of the route therefore matters to interpretation.
Imagine a fictional survey that visits every park first and every commercial street later. If temperature rises during the journey, the resulting difference could exaggerate the contrast between land uses. A reverse-order survey, stationary reference or another suitable design can help investigate the problem. The appropriate correction depends on the measurement and conditions, not on a universal formula applied without checking.
Instrument response time adds another consideration. A sensor does not necessarily equilibrate instantly when it moves from sun to shade or from a sheltered street to an open area. The recorded value may partly reflect the previous environment. Vehicle heat, mounting and airflow can also affect the measurements. These factors should be documented rather than assumed negligible because the device has a digital output.
Juniper Ward could use mobile surveys to identify locations for more detailed follow-up. That is a valuable purpose even if the first campaign does not support a precise long-term ranking. The survey report should state its date, route, time window, instrumentation and how background changes were handled. It should distinguish directly observed points from values interpolated between them.
Community participation can improve the relevance of the route and interpretation. Residents may identify waiting areas, inaccessible shade or unusual local conditions that a remote analyst would miss. Their observations should be integrated respectfully while preserving the distinction between lived experience, instrument readings and causal conclusions. Each contributes something different to the investigation.
Cooling infrastructure needs an operating plan
A newly installed intervention is only the beginning of its useful life. Trees need establishment and care. Shade structures need inspection and repair. Surface treatments can change with wear and soiling. An evaluation that ends on opening day measures a project in its newest condition, not necessarily the conditions people will encounter several summers later.
For Juniper Ward, procurement should identify the intended service rather than only the installed object. A tree count measures planting activity. Surviving canopy, useful shade and access measure different aspects of delivered benefit. A shelter count measures structures; occupied shaded space at relevant hours addresses performance. Both activity and performance can be reported without pretending they are the same metric.
Maintenance budgets should follow the promised service. If a plan relies on vegetation, water availability and long-term care deserve attention. If a structure is damaged, the responsible department needs a process for restoring it. A project whose benefit depends on an unfunded future action should disclose that dependency rather than present the full future benefit as already secured.
There are also practical constraints that should be discussed openly. Underground utilities, pedestrian clearance, sight lines, water needs and local ecology can affect the choice of intervention. Recognising those constraints does not justify leaving exposed places unchanged. It helps compare alternatives that can actually be implemented and sustained.
The strongest design brief asks who will use the space, when they will use it, what measured condition should improve and who will maintain that improvement. This makes later evaluation possible. Without a service definition, a city can meet an installation target while remaining unable to say whether the intended users received meaningful relief.
A plan is an intention until outcomes are measured
Phoenix's official council record shows approval of the Shade Phoenix Plan on November 13, 2024. The item describes policy direction for natural and built shade and prioritising people exposed to heat, including vulnerable populations. That is evidence of an adopted municipal plan. It is not, by itself, a measurement of how much exposure subsequently changed at every site covered by the plan. Phoenix's primary council record
The distinction matters for every city, including fictional Juniper Ward's administration. Plans define goals, responsibilities and priorities. Construction records show what was installed. Monitoring shows how the intervention performed under observed conditions. Use records show who could access it. A credible progress report connects these stages without allowing one to stand in for all the others.
A city might report that a certain number of shelters were installed while acknowledging that seasonal performance measurements are still underway. That is more informative than translating expenditure directly into an unmeasured temperature reduction. It also creates an opportunity to improve later installations based on what the first ones reveal.
The evaluation should preserve less favourable findings. If one shelter provides poor afternoon coverage, the result can guide a redesign. If a surface treatment's effect differs from expectations, the city can reconsider where it is suitable. Treating every outcome as a public-relations threat makes it harder to learn from real conditions.
Public accountability is strongest when people can follow the chain from need to intervention to measured service. The report should state what remains uncertain and what will be checked next. That does not diminish the value of acting. It distinguishes an adaptive programme from a promise whose success is assumed because the ribbon has been cut.
The useful map is attached to a decision
There is no single best urban heat map for every purpose. A land-surface image, a standardised air-temperature network, a pedestrian radiative survey and an indoor monitoring study each reveal different parts of the city. The choice should begin with the decision the information is meant to support.
For Juniper Ward, locating sun-heated surfaces might justify a satellite analysis. Choosing where to place a bus shelter requires attention to people, time and shadow geometry. Evaluating working conditions requires an appropriate occupational assessment. Understanding indoor exposure requires building-level evidence. The same neighbourhood can need all four investigations without any of them being redundant.
A useful public dashboard would keep the quantities separate and explain their connections. It would show dates, measurement or model status, relevant spatial scales and uncertainty. It would avoid converting every result into an undifferentiated feels-like number. It would distinguish a measured improvement from a projected one and a local finding from a citywide average.
The broader lesson is encouraging because it makes interventions more precise. Shade can be valuable even when it barely changes the air. A roof treatment can serve a building objective without claiming a matching pedestrian benefit. Trees can contribute several services while their citywide temperature effect remains different from the experience beneath a canopy. Careful measurement does not reduce those benefits; it allows them to be described and improved honestly.
The temperature on the map becomes useful when someone can answer three questions: what was measured, where and when was it measured, and whose decision does it inform? Until then, a vivid colour is only an invitation to investigate. The goal of urban heat policy is not to make the map look cooler. It is to improve the conditions people actually encounter, with evidence specific enough to show when that improvement has occurred.