Statistical predictions of typical well depths and related spatial diagnostics for two hydrogeologic frameworks through blending cubist and generalized additive statistical models for the Chicot aquifer system, southwestern Louisiana, from the covCHICOTla

This data release provides gridded predictions of “typical” well depth for the hydrogeologically complex Chicot aquifer system, southwestern Louisiana. These predictions were made using an applied statistical prediction modeling workflow named welldepmod_make.R (Asquith and Lindaman, 2025) to estimate groundwater well depths for two custom bundles (classifications) of local aquifer codes for many thousands of wells. Local aquifer codes represent a coding scheme used by the U.S. Geological Survey (USGS) National Water Information System (NWIS). Asquith and Lindaman (2025) provide comprehensive citations, data sources, data filtering decisions, and other details leading to wells being classified into either CHCTLYR1 or CHCTLYR2 classes suitable for statistical production of typical well-depth geospatial products provided in this data release. The CHCTLYR1 (combination of USGS NWIS local aquifer codes 112CHCTU, 11202LC, 112ACFL, 112CHCT, and 112UPTC) represents comparatively "shallow" Chicot aquifer system conditions and effectively includes the entire horizontal expanse of the aquifer. The CHCTLYR2 (combination of USGS NWIS local aquifer codes 112CHCTL,11205LC, and 11207LC) represents comparatively "deep" aquifer conditions and is applicable for only part of the Chicot aquifer system expanse; this area is defined by a computational hull corresponding to the data availability of wells assigned the local aquifer codes associated with CHCTLYR2. The modeling implemented support vector machine classification review to make relatively minor adjustments in the CHCTLYR1 or CHCTLYR2 classification by well using probabilistic estimates using four predictor variables or covariates: (1) horizontal well position, (2) Chicot aquifer system bottom altitudes, (3) Chicot aquifer system top altitudes, and (4) mean depth to water for the period record of the wells. The total number of wells was 8,291 with 7,106 being classified as CHCTLYR1 and 1,185 being classified as CHCTLYR2. Using these wells, the modeling then implemented cubist and generalized additive statistical model learning for prediction of typical well depth. The well-depth predictions were made at the 1-kilometer grid scale using the National Hydrogeologic Grid (Clark and others, 2018). The predictions of typical well depth are intended to provide a first-order and general estimate of groundwater wells for purposes of having continuous gridded data for derivative scientific statistical research including groundwater levels, water quality, and hydrogeologic characteristics. Lastly, the computational hull for CHCTLYR2 also is provided for cartographic purposes and is based substantially on heuristic decisions made during operation of the script welldepmod_make.R (Asquith and Lindaman, 2025); this hull might have usefulness in derivative scientific software endeavors.
 
Note: The authors are aware that some commercial geospatial software report warnings or encounter difficulties related to the coordinate reference system inherited from Clark and others (2018) digital files within the well-depth model workflow as it produced the geopackage-formatted (.gpkg) polygon representation of the computational hull. The authors have tested reading the geopackage file into Python, R, and other open-source software and writing or exporting said geospatial data into the so-called "shapefile" format (.shp) for which use of this format is expected to abate the aforementioned warnings and difficulties in some commercial software.