ABSTRACT:

The source is a repository of Spatially Explicit Estimate of Tile Drainage (SEETileDrain) products across the US Midwest in 2017 at a 30-m resolution. It includes the binary classification map (tile and non-tile), tile probability (how likely a grid cell is tile-drained). The Python scripts to generate the PAW layers and the R scripts (see also: https://github.com/LuwenWan/SEETileDrain_MidWest) to select variables, implement the random forest model and visualize the figures, are also available.

In this work, we developed a machine learning model using 31 satellite-derived and environmental variables and trained with 60,938 tile and non-tile ground truth points within the Google Earth Engine cloud computing platform. The results show that our model achieved good accuracy, with 96 % of the points correctly classified and an F1 score of 0.90. When the tile drainage areas are aggregated to the county scale, it agrees well (R-squared = 0.68) with the reported area from the 2017 Ag Census. The product, SEETileDrain (Spatially Explicit Estimate of Tile Drainage), is described in full detail in the manuscript and the supporting information of Wan et al. (2024). If needed, copies of the tile drainage product can be requested from the corresponding author at luven.wan@gmail.com.

Preferred citation:
L. Wan, A.D. Kendall, J. Rapp, D.W. Hyndman. 2024. Mapping agricultural tile drainage in the US Midwest using explainable random forest machine learning and satellite imagery, Science of the Total Environment. https://doi.org/10.1016/j.scitotenv.2024.175283

ABSTRACT:

SENSEflux-USGLB, the Spatially Explicit Nutrient Source Estimate and Flux for the United States Great Lakes Basin estimates total annual nitrogen and phosphorus loads from the US Great Lakes Basin to the coastline, as well as sources and pathways at 120 meter resolution for an average year during 2008 - 2015 period (ca. 2010). The SENSEflux model uses a GIS and mass balance approach to simulate nutrient fate and transport from point and nonpoint sources across the landscape through rivers to lakes and wetlands. It includes four components: (1) nutrient applications, (2) in situ losses, (3) basin attenuation through surface and subsurface pathways, and (4) stream and lake attenuation. This resource includes 120-meter maps of nitrogen and phosphorus loads, mass balance components (total applied nutrient, crop harvest, basin loss, river uptake, soil and groundwater storage), sources (atmospheric deposition, chemical agricultural fertilizer, chemical nonagricultural fertilizer, manure, septic tanks, nitrogen fixation from legumes, and point sources) and pathways (overland flow, tile drainage, groundwater, septic plumes, point) along with corresponding watershed summaries at the Hydrologic Unit Code 12 (HUC12) and HUC8 levels, as defined in the USGS 2014 Watershed Boundary Dataset. Sources are separated by subsurface (groundwater flow and septic plumes within groundwater) pathway and surface (overland and tile fields) pathway, so total nutrient delivery through a specific nutrient source can be computed as the sum of each source (i.e., gQAgComm + sQAgComm) through the pathways if applicable. Watershed summaries (kg/day) derived from those 120-meter maps (kg/day/cell) using the zonal statistics method with the 'SUM' function. SENSEflux-USGLB is described in full detail in the manuscript and supporting information of Wan et al. (2023) "Important Role of Overland Flows and Tile Field Pathways in Nutrient Transport in Environmental Science & Technology. https://doi.org/10.1021/acs.est.3c03741.

ABSTRACT:

SENSEflux-USGLB, the Spatially Explicit Nutrient Source Estimate and Flux for the United States Great Lakes Basin estimates total annual nitrogen and phosphorus loads from the US Great Lakes Basin to the coastline, as well as sources and pathways at 120 meter resolution for an average year during 2008 - 2015 period (ca. 2010). The SENSEflux model uses a GIS and mass balance approach to simulate nutrient fate and transport from point and nonpoint sources across the landscape through rivers to lakes and wetlands. It includes four components: (1) nutrient applications, (2) in situ losses, (3) basin attenuation through surface and subsurface pathways, and (4) stream and lake attenuation. This resource includes 120-meter maps of nitrogen and phosphorus loads, mass balance components (total applied nutrient, crop harvest, basin loss, river uptake, soil and groundwater storage), sources (atmospheric deposition, chemical agricultural fertilizer, chemical nonagricultural fertilizer, manure, septic tanks, nitrogen fixation from legumes, and point sources) and pathways (overland flow, tile drainage, groundwater, septic plumes, point) along with corresponding watershed summaries at the Hydrologic Unit Code 12 (HUC12) and HUC8 levels, as defined in the USGS 2014 Watershed Boundary Dataset. Sources are separated by subsurface (groundwater flow and septic plumes within groundwater) pathway and surface (overland and tile fields) pathway, so total nutrient delivery through a specific nutrient source can be computed as the sum of each source (i.e., gQAgComm + sQAgComm) through the pathways if applicable. Watershed summaries (kg/day) derived from those 120-meter maps (kg/day/cell) using the zonal statistics method with the 'SUM' function. SENSEflux-USGLB is described in full detail in the manuscript and supporting information of Wan et al. (2023) "Important Role of Overland Flows and Tile Field Pathways in Nutrient Transport in Environmental Science & Technology. https://doi.org/10.1021/acs.est.3c03741.