2026/03/10 by Akira Byrne, David P. Huber, Dylan Thompson +2 · 1 voice
Environmental Science · #Hydrology and Watershed Management Studies #Soil Geostatistics and Mapping #Water Quality and Resources Studies
paper · doi:10.18122/geo_data.10.boisestate
openalex publication_date 2026/03/10 · openalex created_date 2026/03/11 · openalex updated_date 2026/07/31
Soil depth is critical to ecosystem productivity, acting as a reservoir for water, nutrients, and carbon. Deeper soils enhance water holding capacity to withstand droughts in native and agricultural systems, and provide more volume for long-term carbon sequestration, directly influencing soil health and climate regulation. This dataset scraped public well log (AKA well driller’s report) data from the Idaho Department of Water Resources (IDWR) to create a soil depth map across the Magic Valley in south-central Idaho, USA. Soils in this region are predominantly loess-derived silt-loams, underlain by undulating beds of basalt on top of deep rhyolite. Starting in the early 1900s, the area has been intensively irrigated and transformed into a hub of large-scale crop and dairy production. The greater Magic Valley region includes > 37,000 households, with ~28% or 10,500 households in rural areas. The IDWR hosts well logs for every groundwater well in Idaho. The well logs contain lithologic data including soil depth (i.e., the soil-bedrock interface). We manually extracted soil depths from ~8,000 well sites within the intensive agricultural area in Magic Valley. The boundaries of the study area included the Snake River Canyon as the northern watershed boundary, the High Line Canal as the southern boundary, the Salmon Falls Creek Canyon as the western boundary, and the HUC 12 subwatersheds near Kimberly as the eastern boundary. The study area is 931 km2. Well installations and logs ranged from 1939 to 2024, and varied in the descriptive rigor of the soil depth. Soil depth data was manually flagged as low, medium, and high confidence in the accuracy of the soil depth estimate, leaving a dataset of ~5,700 high quality soil depths. A usable soil depth map of the study region was generated using a standard kriging interpolation method with a gaussian semivariogram model. Edge effects were minimized by including a 1.5 km buffer around the area of interest. Soil depths ranged from < 0.6 m (2 ft) to > 8.9 m (29.5 ft). Ultimately, well logs provided a usable and high quality public data source for mapping soil depth in a dryland region. Soil depth (i.e., control volume) limits landscape-scale water, nutrient, and carbon storage and losses, and may be an important factor when predicting system productivity and variability with change in climate and/or land use.