Using imaging spectroscopy and elevation in machine learning to estimate soil salinity in intermittently tidal wetlands

German Silva; Dar Roberts; Kristin B. Byrd; Dana Chadwick; Ian Walker; Jennifer King

doi:10.1002/ecs2.70356

Home/Archive/USGS

USGS · 70269614

Using imaging spectroscopy and elevation in machine learning to estimate soil salinity in intermittently tidal wetlands

German Silva·Dar Roberts·Kristin B. Byrd·Dana Chadwick·Ian Walker·Jennifer King

View original source ↗Export metadata ↓

Abstract

Coastal soil salinization patterns are changing due to drought, sea level rise (SLR), and changing freshwater inflow. These changes are expected to impact coastal wetland plant health and ecosystem function, such as changes to biomass and productivity. These impacts have led to greater interest in how we monitor soil salinization across spatial and temporal scales. Remote sensing is a promising tool for estimating soil salinity at the spatial scales required for decision making by land managers. However, the development of a remote sensing estimation approach for wetland soil salinity must account for two factors: (1) the high spatial and temporal heterogeneity of coastal wetlands and (2) the fact that soil salinity is the result of multiple historical land use, hydrological, and geomorphic processes. In spring 2022, a combined airborne-field campaign, known as SHIFT, collected a weekly time series of airborne visible to shortwave infrared (VSWIR) image spectroscopy data. This dataset provides a unique opportunity to assess the application of fine spatial (5 m) and temporal (weekly) resolution VSWIR data to estimate root zone soil salinity; when combined with environmental variables such as elevation, these data can account for some of these factors. In this study, we utilized VSWIR and elevation datasets in a random forest regression to predict and map soil salinity in an intermittently tidal estuary, Devereux Slough, located in Santa Barbara County, California. The final model combined spectral indices with elevation to better capture soil salinity dynamics despite lower correlation ( r = 0.85) than solely using elevation ( r = 0.92). This research demonstrates the utility of remote sensing datasets, namely, elevation and the modified Anthocyanin Reflectance Index (mARI), for predicting root zone soil salinity in intermittently tidal coastal wetlands. These findings are an important step in advancing coastal remote sensing by creating a gridded salinity dataset that can be used for salinity monitoring and other coastal applications, such as modeling change in vegetation communities or ecosystems facing the impacts of climatic variability and change.

Explore related subjects

Ecosphere Article

Source-reported bounding extent: 34.406950669793815° to 34.4236991476653° latitude; -119.88462619707985° to -119.86741425943123° longitude. This indicates report coverage, not an exact sampling location. View area on OpenStreetMap.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

German Silva, Dar Roberts, Kristin B. Byrd, Dana Chadwick, Ian Walker, Jennifer King. 2025-08-05. Using imaging spectroscopy and elevation in machine learning to estimate soil salinity in intermittently tidal wetlands. https://doi.org/10.1002/ecs2.70356

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

KEEP EXPLORING

Related USGS reports

USGS2026

Fire regime shift reveals 17th century socio-ecological change in the Rio Grande del Norte National Monument, New Mexico

Humans have influenced ecosystems for millennia, and social changes can have notable ecological effects. Recent human influence on ecosystems is exemplified by the collapse of fire regimes circa 1900 in the western United States due to increased grazing, fire suppression, and declining Indigenous fire use. Climate can also drive fire regime changes, which is reflected by increasing area burned across the West in recent decades associated with warming and increased aridity. In the southwestern United States during the mid-1600s, large social changes accompanied Spanish colonization. These changes included increases in domestic livestock (e.g., sheep and horses), declines of Indigenous populations and cultural burning, and exploitation of Puebloans that led to the 1680 Pueblo Revolt, an Indigenous uprising that expelled Spanish colonizers for 12 years. The ecological effects of these large social changes are not well studied. The goal of this study was to use tree-ring fire histories to investigate fire dynamics of a sagebrush-dominated landscape in the Rio Grande del Norte National Monument (RGDN) in northern New Mexico, including the effects of historical social and climatic changes. We reconstructed fire history on four forested mountains that rise above the sagebrush/grass plateau. We asked: (1) What was the pre-1900 fire frequency, season, and fire–climate relationship in the RGDN? (2) Was there evidence for fire spread in the sagebrush/grass flats between the mountains? and (3) Did regional social or climate changes influence fire regimes? From 1400 to 1679, we found frequent fire within forests (8–12-year median fire intervals [mfi]) and potential evidence of spreading fire through the sagebrush/grass ecosystem based on greater-than-expected intermountain fire synchrony. However, beginning in the mid-1600s and continuing after the Pueblo Revolt of 1680, there was (1) a significant decline in fire frequency, including 38–166-year mfi and no fire for >350 years at two sites; and (2) only one synchronous fire between mountains. During this 17th century period of socio-ecological change there was no persistent shift in climate. We propose that reduced Indigenous ignitions and increased domestic grazing, which reduced herbaceous fuels and fire spread, are likely causes of the extraordinary early and persistent reduction in fire.

New Mexico↗

USGS2026

Climate-driven changes in wetland hydroperiods predict losses in habitat suitability for amphibian breeding

Environmental regimes are shifting with accelerating climate change, disrupting interactions between species and ecosystems shaped by the climate of the pre-industrial era. Species assemblages that depend on a dynamic network of ephemeral and permanent habitats may be especially sensitive to changing climate, whereby shifting environmental regimes may limit species persistence in landscapes in ways not captured by species physiology or coarse habitat metrics alone. Here we forecasted the loss of montane and subalpine wetlands as breeding habitat for three amphibian species in the US Pacific Northwest, Ambystoma gracile , Ambystoma macrodactylum , and Rana cascadae . We combined historical and recent surveys to build occupancy models of current breeding habitat use. We combined these results with existing hydrological models to predict the loss of suitable breeding habitats by the 2080s due to increased wetland drying. We found that species-specific occupancy of wetlands used for reproduction was predicted by both abiotic (e.g., elevation, wetland area) and biotic (e.g., emergent vegetation) site-level characteristics. Among wetlands currently used for reproduction, we estimated that between 12% and 23% could become unsuitable for the aquatic phase of each species by the 2080s, with watershed-scale losses ranging from 0% to 80%. This suggests that wetland habitat loss could limit the persistence of these species in the Pacific Northwest, though impacts are expected to vary across landscapes. More broadly, we demonstrate a novel approach to quantifying species climate change vulnerability that can be used to identify areas of both risk and resilience in changing landscapes and help to prioritize the application of practical conservation solutions in areas of greatest need.

Washington↗

USGS2026

Sex-specific reproductive investment and survival of adult Atlantic salmon kelts across multiple dams

In adult Atlantic salmon, individual variation in migration and movement within rivers during spawning (i.e., reproductive movement, RM) is an important component of overall reproductive investment that may influence subsequent survival and repeat spawning. We test the cost-of-reproduction hypothesis—which states that increased reproductive investment decreases survival—by first quantifying individual RM of landlocked Atlantic salmon and then evaluating how RM, sex, and dams influence outmigration survival. Consistent with sexual selection for male mobility, male RM was ~4 times greater than that of females: they moved further, were more exploratory, visited a tributary more, went further upstream, and spent longer in the river. Results provide mixed support for a cost of RM on outmigration survival: individuals (inclusive of both sexes) that abandoned spawning and migrated downstream (i.e., “fallbacks”) survived 3.5-fold better than those that spawned (supporting an overall cost-of reproduction). However, males (with higher RM) had 2.8 times higher survival than females, and among males there was no effect of RM on survival (not-supporting a cost-of-RM). The most influential factor on survival was dam interactions, which reduced per-kilometer survival 18-fold relative to migration within free-flowing reaches. Documenting male-biased survival was unexpected given the (1) high RM of males and (2) that most prior studies find higher outmigration survival of female Atlantic salmon. We highlight how dams exacerbate the already high costs of reproduction in Atlantic salmon, which has important implications for contemporary evolution of behavioral tradeoffs.

Vermont↗