AI-OCR driven recovery of legacy soil data unlocks decades of buried information at a fraction of the original cost. This work demonstrates scalable methods for rescuing soil data from paper and PDF archives across universities and NRCS offices. Historical pedon records, which originally required $8,000–$10,000 per site to collect and analyze, were digitized and structured…
Poster – 2024 NCSS Regional Conference, Platteville, WI This study demonstrates that a bottom-up, locally enhanced DSM framework can exceed the performance of one of the most detailed conventional soil maps available. The LE-DSM product outperformed SSURGO for 46% of soil property and depth predictions, while SSURGO exceeded top-down DSM products in 75% of comparisons….
Meyer Bohn and Bradley Miller – National Cooperative Soil Survey National Conference June 23-25, 2025 GPT-4 TurboDownload Abstract Advancements in artificial intelligence (AI) and optical character recognition (OCR) technology are rapidly transforming how we process historical soils data. Just two years ago, large-scale soils legacy data rescue required manual entry, but today, AI offers a…
Caner Ferhatoglu and Bradley Miller – North Central Soil Fertility Conference November 16-17, 2022 Caner-Ferhatoglu-Poster-North-Central-Soil-Fertility-Conference-Final-VersionDownload Abstract In this study, the effectiveness of six types of FS methods from four categories (filter, wrapper, embedded, and hybrid) were compared. These FS algorithms chose relevant covariates from a set of 1049 environmental covariates for predicting five soil fertility…
Meyer Bohn and Bradley Miller – Iowa Water Center Conference April 6-8, 2021 Bohn-Dows-Topsoil-Thickness-IWC Abstract Abstract The objectives of this study were to assess spatial predictions of topsoil thickness from models produced from ensemble machine learning algorithms along with assessing the uncertainty estimations associated with those models. Boosting is one example of an ensemble method,…
A statistical comparison of soil properties classified via Digital Hillslope Position and Floodplain algorithm vs. gSSURGO and Iowa physiographic regions.
A new approach for delineating noncontributing areas of watersheds is presented. Maps of soil properties offer indicators of subtle landscape features and hydrologic functions. Our approach is to identify soil map units that reflect upland ponding and then define the sum of their watersheds as the noncontributing area. The decrease in time required to delineate…
Grid, random, and stratified sampling each come with unique advantages and disadvantages.
However, more efficient sampling could be accomplished with a more strategic sampling design.
The map produced by William Smith in 1815 is recognized as the first geological map of Britain, but it also represents a benchmark in cartography for the Earth sciences. Following the agrogeology concepts of the time, the title claims that by mapping the geologic strata, the varieties of soil are also shown. The new opportunity…
Native prairie vegetation strips on contours (STRIPS) are a relatively new conservation strategy which bring a host of ecosystem services. They also add a degree of complexity to what would otherwise be a simplified landscape. This study analyzes the impact of the presence of the prairie strips on surrounding soil properties.
