Abstract
Abstract
High-resolution measurements of near-surface temperature are essential for understanding environmental responses to extreme weather events, yet are rarely available from dense, regularly spaced in situ measurement grids. Here, we analyse temperature measurements derived from a dense network of over 3,000 nodal seismometers deployed across a 6 km² area in North Yorkshire, UK, during the July 2022 heatwave, using data from embedded microcontroller sensors. The dataset comprises over 80 million temperature measurements at 100 s intervals over 34 days.
Microcontroller derived temperature measurements were statistically corrected using co-located meteorological observations to account for systematic biases introduced by the thermal response of the sensor housing to closer approximate near-surface grass temperatures. The corrected dataset reveals strong spatial heterogeneity in near-surface temperatures, with mean differences of up to 3.6 °C over distances of 100–200 m and extreme contrasts exceeding 17 °C during peak heating conditions. Spatial variability is strongly modulated by meteorological forcing, with daytime heatwave conditions exhibiting the greatest heterogeneity, while nighttime periods show substantial homogenisation. Systematic but modest differences are observed between land-use classes (~0.1 °C), expressed primarily through variations in persistence and extremes rather than mean temperature alone.
The results demonstrate the potential of repurposing embedded microcontroller temperatures within geophysical instruments to provide dense environmental observations, improving the characterisation of microscale temperature variability and supporting applications in ecosystem monitoring, the built environment and critical infrastructure.