This dataset contains the final water level records (10,132) that met all threshold criteria and had absolute residual (ABS_PSEUDO_LEV) of less than 20 feet. The threshold of 20 feet was selected based on water level changes of the U.S. Geological Survey groundwater site 363551090152801 Qulin in the Mississippi River Valley alluvial aquifer. This final dataset is considered a best representation of the true water levels in the aquifer.

An objective review (Asquith and others, 2018 and 2020) of the distribution of the first two significant figures of a water-level measurement (depth below land surface) was done on the 10,295 measurements (one per well) that met the threshold criteria. The purpose of this review was to ascertain the degree to which substantial rounding of values might exist in the dataset. It was evident that the dataset has a large number of values rounded to the nearest integer foot with a tendency for more rounding towards even integers. For values between 10 and 99 feet, there is a large number of values rounded to the even 10 feet and for values less than about 35 feet, there are an excessive number of values rounded the nearest. There are also numerous values of 12 and 14. This suggests that considerable estimation or rounding of water levels have been made probably by use of apparatus other than graduate tapes. Systematic review of original data sources is not possible and insufficient metadata exist for a manual of each value in the data set. However, the database is large and offers an opportunity for data mining and machine learning to foster further review. A large, objective, technically-demanding, and rigorous spatial-temporal review of the water-level data, expressed in altitude, was made for the 10,295 water level records comprising this data release.

An objective review (Asquith and others, 2018 and 2020) of the distribution of the first two significant figures of a water-level measurement (depth below land surface) was done on the 10,295 measurements (one per well) that met the threshold criteria. The purpose of this review was to ascertain the degree to which substantial rounding of values might exist in the dataset. It was evident that the dataset has a large number of values rounded to the nearest integer foot with a tendency for more rounding towards even integers. For values between 10 and 99 feet, there is a large number of values rounded to the even 10 feet and for values less than about 35 feet, there are an excessive number of values rounded the nearest. There are also numerous values of 12 and 14. This suggests that considerable estimation or rounding of water levels have been made probably by use of apparatus other than graduate tapes. Systematic review of original data sources is not possible and insufficient metadata exist for a manual of each value in the data set. However, the database is large and offers an opportunity for data mining and machine learning to foster further review. A large, objective, technically-demanding, and rigorous spatial-temporal review of the water-level data, expressed in altitude, was made for the 10,295 water level records comprising this data release.

This dataset contain the final 6,298 water-level records that met all threshold criteria and had absolute residuals (ABS_PSEUDO_LEVRES) not exceeding 20 feet and exceeding the 99th percentile. This final dataset is considered a best representation of the true water levels in the aquifer.

A potentiometric-surface map represents the altitude at which water would stand in tightly cased wells completed at any location within the study area aquifer. Using the altitude of water levels measured in the study area, the potentiometric-surface map depicts points of equal altitude with contours denoting a given water-level altitude calculated by subtracting the water level measured from the land-surface elevation (National Geodetic Vertical Datum of 1929). The contour lines were created using computer-based program ArcGIS with an interval of 10 feet. The direction of water flow from areas of high elevation to low elevation can be interpreted using potentiometric-surface maps and areas of decreased groundwater levels can be identified. The 2014 potentiometric-surface map shows ten total cones of depressions: two large depressions, five small depressions, and three areas of decreased water levels. As with the 2010 potentiometric-surface map, one large depression begins in southeastern Arkansas County, near the Arkansas and Desha County line, and extends north into Prairie County, west into Lonoke County, and east into the western-most part of Monroe County. Even though the center of the depression had deepened in 2010, the area of the cone in Arkansas County within the southeastern half of the depression had not expanded horizontally. The analysis of the 2014 potentiometric-surface map suggests no horizontal expansion in this area. The additional GIS shapefiles were used to depicts the western extent of the Mississippi River alluvial aquifer in eastern Arkansas on plates 1, 2, and 3 in Rodgers and Whaling (2020).

A potentiometric-surface map represents the altitude at which water would stand in tightly cased wells completed at any location within the study area aquifer. Using the altitude of water levels measured in the study area, the potentiometric-surface map depicts points of equal altitude with contours denoting a given water-level altitude calculated by subtracting the water level measured from the land-surface elevation (National Geodetic Vertical Datum of 1929). The contour lines were created using computer-based program ArcGIS with an interval of 10 feet. The direction of water flow from areas of high elevation to low elevation can be interpreted using potentiometric-surface maps and areas of decreased groundwater levels can be identified. The 2014 potentiometric-surface map shows ten total cones of depressions: two large depressions, five small depressions, and three areas of decreased water levels. As with the 2010 potentiometric-surface map, one large depression begins in southeastern Arkansas County, near the Arkansas and Desha County line, and extends north into Prairie County, west into Lonoke County, and east into the western-most part of Monroe County. Even though the center of the depression had deepened in 2010, the area of the cone in Arkansas County within the southeastern half of the depression had not expanded horizontally. The analysis of the 2014 potentiometric-surface map suggests no horizontal expansion in this area. The additional GIS shapefiles were used to depicts the western extent of the Mississippi River alluvial aquifer in eastern Arkansas on plates 1, 2, and 3 in Rodgers and Whaling (2020).

This dataset includes grid node locations at which groundwater altitudes were estimated for April 10, 1980 and April 10, 2016 using the general additive model (GAM) and groundwater observations from the USGS NWIS web interface.