For the past 18 months, TerraFuture has operated a growing network of 312 low-cost environmental sensors deployed across Portland's diverse neighborhoods. What started as a pilot project with 40 sensors has evolved into one of the most granular urban heat monitoring systems in the Pacific Northwest. The data we have gathered tells a story that is both striking and actionable.
The Temperature Gap Is Wider Than Expected
Our sensor data reveals that on a typical summer afternoon, temperature differentials between Portland's most and least tree-covered neighborhoods can reach 11.4 degrees Fahrenheit. That number is not an outlier from a single heat wave event. It represents the median peak differential recorded across 94 days of summer monitoring in 2025.
The neighborhoods registering the highest sustained temperatures correlate strongly with historically redlined districts. Areas in East Portland, where tree canopy coverage drops below 15 percent, consistently recorded temperatures 8 to 12 degrees above those in the West Hills, where canopy coverage exceeds 40 percent.
Methodology: Hyperlocal Data at Scale
Each of our sensor nodes measures ambient temperature, relative humidity, particulate matter (PM2.5), and surface albedo at five-minute intervals. The units are calibrated against NOAA reference stations quarterly, and our data pipeline applies automated quality-control algorithms to flag and correct drift.
We partnered with Portland State University's Urban Studies department to validate our readings against satellite-derived land surface temperature data. The correlation between our ground-level network and Landsat thermal imagery sits at r = 0.91, giving us high confidence in the spatial patterns we observe.
The convergence of ground-level sensor data and satellite imagery creates a picture of urban heat that is difficult to ignore. These are not abstract climate projections. They are measurements of today's reality.
Key Findings
Nighttime cooling disparities matter most for health. While daytime temperature differences receive the most attention, our data shows that nighttime cooling gaps are where the real health risk concentrates. Neighborhoods with less than 20 percent canopy coverage cool an average of 4.2 degrees less overnight than heavily treed areas. This sustained heat exposure during sleeping hours is directly linked to increased heat-related illness.
Paved surface area is the strongest single predictor. Through regression analysis of 14 land-cover variables, impervious surface percentage emerged as the most significant predictor of elevated temperatures, accounting for 67 percent of the variance in our model. Tree canopy was the second strongest factor at 21 percent.
Micro-interventions show measurable results. In three neighborhoods where TerraFuture installed reflective surface treatments and small-scale tree plantings in 2024, we measured a statistically significant 1.8-degree reduction in peak afternoon temperatures by the end of the 2025 summer. This effect will likely grow as saplings mature.
What This Means for Portland
Our data supports what environmental justice advocates have long argued: urban heat is not distributed equally, and its burdens fall disproportionately on communities of color and low-income neighborhoods. The sensor network gives us the evidence base to move from anecdotal observation to targeted intervention.
TerraFuture is sharing this dataset publicly through our Open Climate Data Portal, and we have begun working with the Portland Bureau of Planning and Sustainability to integrate our findings into the city's updated Climate Action Plan.
What Comes Next
In 2026, we are expanding the network to 500 sensors and extending coverage to Gresham and Milwaukie. We are also developing a real-time heat vulnerability dashboard that will combine our sensor data with demographic and health data to help emergency managers direct cooling resources during heat events.
The data is clear. The question now is whether our policies and investments will match the urgency that the numbers demand.
