SatCHM (Satellite Canopy Height Model): Leveraging deep learning for site-specific sub-meter canopy height predictions

Mitchell, M.; Abolt, C.; Crennen, Z.; Marcato, A.; Atchley, A.

doi:10.64898/2026.08.25.728853

Home/Archive/bioRxiv

bioRxiv · 10.64898/2026.08.25.728853

SatCHM (Satellite Canopy Height Model): Leveraging deep learning for site-specific sub-meter canopy height predictions

Mitchell, M.·Abolt, C.·Crennen, Z.·Marcato, A.·Atchley, A.

View original source ↗Export metadata ↓

Abstract

High-resolution monitoring of forest structure and productivity is essential for effective natural resource management. However, monitoring approaches such as field-based forest inventories or extensive lidar campaigns are costly, time-intensive, and spatially limited. Therefore, inexpensive and accessible methods are needed. SatCHM (Satellite Canopy Height Model) was developed to be an accessible and open-source tool for researchers, allowing for site-specific and temporally flexible predictions of canopy height with limited computational resources. SatCHM requires four inputs: panchromatic satellite imagery, solar and sensor angle metadata of satellite imagery, digital elevation models (DEMs), and lidar-produced CHMs for an area of interest. After SatCHM pre-processes inputs, data is loaded into a collection of convolutional neural networks (CNNs) for image-to-image regression. This ensemble cooperates to yield high-resolution predictions (up to 0.5-meter) of three-dimensional tree structure with discernible tree crowns across a broader defined area of interest. After calculating the mean absolute error for each prediction output, the median of these mean absolute errors was 6.06 meters.

Explore related subjects

ecology Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Mitchell, M., Abolt, C., Crennen, Z., Marcato, A., Atchley, A.. 2026-08-27. SatCHM (Satellite Canopy Height Model): Leveraging deep learning for site-specific sub-meter canopy height predictions. https://doi.org/10.64898/2026.08.25.728853

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Wildfires drive trade-offs in ammonia-oxidizing groups and promote denitrification to increase soil emissions of nitric oxide (NO) and nitrous oxide (N2O) in California chaparral

Wildfires can promote trade-offs in soil nitrifier and denitrifier communities that affect post-fire nitrogen (N) cycling and emissions of nitric oxide (NO), nitrous oxide (N2O), and dinitrogen (N2). For example, by increasing soil pH and ammonium (NH4+), wildfires could increase the abundance of ammonia-oxidizing bacteria (AOB) over archaea (AOA). Because AOB and AOA process N differently, nitrifier trade-offs may affect N emissions and nitrate (NO3-) supply with downstream effects on denitrifier activity, leading us to ask: Do trade-offs between AOA and AOB abundance and shifts in denitrifying processes influence N emissions over time after wildfire? We selectively inhibited AOA and AOB communities from soil collected over four seasonal time points one year before and after a chaparral wildfire and used stable isotopes to parse denitrifier contributions to N2O emissions. Over one year after the wildfire, soil pH increased from 6.1 to 7.0, soil extractable NH4+ increased 30-fold, NO3- increased 5-fold, and NO2- increased 9-fold. AOB amoA gene copy numbers increased 20-fold one year after fire, while AOA abundance remained unchanged. Post-fire soil NO emissions increased 63-fold over one year, with varied contributions from all nitrifier groups. Soil N2O emissions peaked eight months after fire (2271 {+/-} 634 ng N2O-N g-1 soil), with increased contributions from AOA. Wildfire increased {delta}15NSPN2O and {delta}15NbulkN2O values, suggesting increased N2O reduction to N2. Overall, wildfire increased AOB abundance relative to AOA, promoting nitrification activity and providing intermediates to denitrifiers to increase emissions of NO and N2O for up to one year after fire.

ecology↗

bioRxiv2026

Global urban forest cover and nature access is insufficient for human wellbeing and worsening

Target 12 of the Global Biodiversity Framework urges governments to enhance human access to nature to improve human wellbeing. Here, we study nature access using several metrics for a globally representative sample of 140 large Functional Urban Areas, examining variation among regions and over time. We find most urbanites globally do not reach common benchmarks for urban forest cover and nature access. Globally, 3.3 billion people (76% of urbanites) live in urban areas that do not have at least 30% tree canopy cover. Moreover, 1.1 billion people (23% of urbanites) are farther than 1 km from a park as identified in Open Street Map, and 1.5 billion people (36% of all urbanites) are more than 1 km from a green patch as defined by land cover. Over time, urban nature access has declined. Between 1992 and 2020, most urban residents had an increase in the distance to green patches (63% of urbanites increased, average increase 680 m). Similarly, between 1992 and 2020, 31% of urbanites had a decrease in percent natural cover, while only 1.3% had an increase. We show that the variation among cities or regions in the amount of nature access depends on the metric used and the context. On average, mesic climates have more tree cover than do arid climates, and cities with lower GDP or higher population density tend to have lower natural share than other cities. Our results suggest that global urban nature access is insufficient for supporting human health and worsening over time.

ecology↗

bioRxiv2026

Exploring soil microbes' potential to assess Atlantic Forest restoration trajectories

Soil microbe communities are key indicators of ecosystem recovery, yet their integration into operational restoration monitoring remains limited. We evaluated the potential for benchmarking early Atlantic Forest restoration with reference site-based comparisons using bacteria and fungi. Samples were collected from three geographic clusters and spanned a pasture-restoration-forest land-use gradient. Using metabarcoding, we assessed whether soil microbe community diversity, structure, and functional profiles respond to restoration progress considering influences from spatial arrangement and/or successional stage (determined using NDVI). We found that OTU alpha diversity was not explained by either predictor, whereas the Shannon diversity of bacterial functions declined significantly with increasing NDVI. Bacterial and fungal OTU composition (Bray-Curtis) responded significantly to both NDVI and geographic cluster, as did microbe functional profiles, although only NDVI proved both marginally and conditionally significant. We also found that soil chemical parameters were primarily structured by geographic cluster but not by NDVI, suggesting that many abiotic edaphic conditions reflect soil history rather than current vegetation status. These findings suggest that both microbe taxonomic and functional metrics track revegetation progress and validate eDNA as a scalable, sensitive restoration monitoring framework. Importantly, our results also support the use of reference site-based monitoring for Atlantic Forest restoration.

ecology↗