Sharon L. Qi Jason J. Gurdak 2006 raster digital data U.S. Geological Survey Data Series USGS DS 192 Denver, CO U.S. Geological Survey http://water.usgs.gov/lookup/getspatial?ds192_hp_npctprob This raster data set represents the percentage of probability of nonpoint-source nitrate contamination (greater than the proposed background concentration of 4 milligrams per liter (mg/L) as N) of recently (defined as less than 50 years) recharged ground water in the High Plains aquifer of the United States. The High Plains aquifer covers approximately 175,000 square miles in eight States; Colorado, Kansas, Nebraska, New Mexico, Oklahoma, South Dakota, Texas, and Wyoming. Elevated nitrate concentrations above the background concentration have been detected in recently recharged (less than 50 years) ground water in the High Plains aquifer. This data set is derived from empirical models developed using multivariate logistic regression to evaluate the vulnerability of the High Plains aquifer to nitrate contamination from nonpoint sources. This data set was generated in a geographic information system from these models and represents the spatial extent of vulnerability of nitrate contamination greater than 4 mg/L across the aquifer. The purpose of the data set is to identify the predicted probability of detecting a concentration of nitrate greater than 4 mg/L as N in recently (less than 50 years) recharged ground water of the High Plains regional aquifer. This raster data set is associated with the U.S. Geological Survey interpretive report SIR2006-5050 by Gurdak and Qi (2006) that describes the development of the empirical models represented by this raster data set. Two multivariate logistic regression models of vulnerability were found to have the most statistical significance, best model fit, and predictive ability. These models, one representing the Northern High Plains and one representing the combined Central and Southern High Plains, indicate that vulnerability to nitrate contamination greater than 4 mg/L is best explained by nonirrigated and irrigated agricultural lands, organic-matter content of soil, depth to the regional water table, and percentage of clay content in the unsaturated zone. The northern, central, and southern regions of the High Plains aquifer are defined by the two narrowings of the aquifer boundary, one near Sharon Springs, Kans., near the border of Kansas and Colorado and the other near the city of Amarillo, Tex. See Gurdak and Qi (2006) for a location map. The regions are referred to in this documentation as NHP (Northern High Plains region), CHP (Central High Plains region), and SHP (Southern High Plains region). This electronic report was subjected to the same review standard that applies to all U.S. Geological Survey reports. Technical reviewers were asked to check the logical consistency, attributes, raster statistics, projection and geographic extent. The technical reviewers checked the metadata for completeness and accuracy. REFERENCES: Cederstrand, J.R., and Becker, M.F., 1998, Digital map of hydraulic conductivity for the High Plains aquifer in parts of Colorado, Kansas, Nebraska, New Mexico, Oklahoma, South Dakota, Texas, and Wyoming: U.S. Geological Survey Open-File Report 98-548, accessed March 2005 at URL http://water.usgs.gov/GIS/metadata/usgswrd/XML/ofr98-548.xml Gurdak, J.J., and Qi, S.L., 2006, Vulnerability of recently recharged ground water in the High Plains regional aquifer to nitrate contamination: U.S. Geological Survey Scientific Investigations Report 2006-5050, 39p. Gutentag, E.D., Heimes, F.J., Krothe, N.C., Luckey, R.L., and J.B. Weeks, 1984, Geohydrology of the High Plains aquifer in parts of Colorado, Kansas, Nebraska, New Mexico, Oklahoma, South Dakota, Texas, and Wyoming: U.S. Geological Survey Professional Paper 1400-B, 63 p. Lorenz, D.L., Goldstein, R.M., Cowdery, T.K., and Stoner, J.D., 2003, Comparison of two methods for delineating land use near monitoring wells used for assessing quality of shallow ground water: U.S. Geological Survey Water-Resources Investigations Report 2003-4067, 13 p., accessed December 2005 at URL http://pubs.usgs.gov/wri/wri03-4067/ Qi, S.L., Konduris, A., Litke, D.W., and Dupree, J., 2002, Classification of irrigated land using satellite imagery, the High Plains aquifer, nominal data 1992: U.S. Geological Survey Water-Resources Investigations Report 2002-4236, 31 p., accessed December 2005 at URL http://pubs.usgs.gov/wri/wrir02-4236/pdf/wri02-4236.pdf Qi, S.L., Konduris, A., Litke, D.W., and Dupree, J., 2002, Location of irrigated land classified by satellite imagery - High Plains area, nominal date 1992: U.S. Geological Survey Open-File Report 2002-441, accessed January 2005 at URL http://co.water.usgs.gov/nawqa/hpgw/GIS.html U.S. Department of Agriculture, 1991, State Soil Geographic (STATSGO) database: U.S. Department of Agriculture, Soil Conservation Service, Miscellaneous Publication 1492, 88 p. U.S. Geological Survey, 1992, National Land Characteristics Data Set (NLCD) , nominal 1992, accessed November 3 2005 at URL http://edc.usgs.gov/products/landcover/nlcd.html 2006 publication date Complete None planned -106.016458 -96.257462 43.807085 31.652762 None Vulnerability Ground water Nitrate contamination None High Plains aquifer Colorado Kansas Nebraska New Mexico Oklahoma South Dakota Texas Wyoming Great Plains region western U.S. None This map does not show actual contamination of recently recharged ground water; rather, the map depicts areas of the aquifer that have the predicted likelihood of recently recharged ground water with nitrate concentrations that exceed the proposed background concentrations of 4 mg/L as N. Although the probability maps show predictions of nitrate detection as a percent probability, there is inherent uncertainty within these predictions that is not shown in the probability map. This data set should not be used for interpretation at less than 500-meter resolution. U.S. Geological Survey Director, Colorado Water Science Center mailing address

Box 25046, MS 415, Denver Federal Center

Lakewood CO 80225 USA (303) 236-4882 (303) 236-4912 GS-W-COden DC@usgs.gov Microsoft Windows XP Version 5.1 (Build 2600) Service Pack 2; ESRI ArcCatalog 9.0.0.535 See Entity Attribute information. Not applicable for raster data. Data set is complete for the whole High Plains area where all of the input data sets were present at a given location. All of the input data sets did not have the exact same extent for the High Plains aquifer boundary; therefore, the extent of this data set is the intersection of all the input data sets and may not match exactly with published versions of the High Plains aquifer boundary. Data gaps within the High Plains boundary indicate areas where the aquifer is not present or where there was no data in any one of the five input data sets used to calculate this data set. The data set has a horizontal resolution of 500 meters. None Sharon L. Qi, Alexandria Konduris, David W. Litke, and Jean Dupree 2002 raster digital data U.S. Geological Survey Open-File Report OFR 2002-441 Lakewood, CO U.S. Geological Survey http://co.water.usgs.gov/nawqa/hpgw/GIS.html Raster data 1992 ground condition U.S. Geological Survey OFR 2002-441 Model input variable. Sharon L. Qi 2004 raster digital data Unpublished data at U.S. Geological Survey. Raster data set 1992 ground condition Nonirrigated land Model input variable G.E. Schwarz R.B. Alexander 19950901 vector digital data Open-File Report U.S. Geological Survey OFR 95-449 Reston, VA U.S. Geological Survey http://water.usgs.gov/lookup/getspatial?ussoils 250000 Vector data 19950901 publication date STATSGO Model input variable (percent organic material) Sharon L. Qi 2004 raster digital data Unpublished data at U.S. Geological Survey. Depth to water data are from year 2000. Raster data set 2004 ground condition Depth to water Model input variable Sharon L. Qi 2004 raster digital data Unpublished data at U.S. Geological Survey Raster data set 2004 ground condition Clay percentage in unsaturated zone Model input variable The GIS was used to extract anthropogenic and hydrogeologic information and summarize the data from 31 data layers (see Gurdak and Qi, 2006, table 5). The data were extracted at well locations where water-quality data of nitrate concentrations were available so that the logistic regression analysis could identify the layers (variables) that were significant in explaining the observed nitrate concentrations. One of the variables found to be significant in the model and used to calculate the probability map was the amount of irrigated land upgradient from a well location. The raster map of the location of irrigated lands in the High Plains classified by Qi and others (2002) was used to calculate the amount of irrigated land within a 90-degree sector oriented upgradient (180 degrees from the regional ground-water flow direction) from each well. The 90-degree sector represents a simulated contributing area for each well and was shown to be a more statistically significant way to characterize data around the well at the surface (Lorenz and others, 2003). The radius of each 90-degree sector was determined by the range of hydraulic conductivity (K) values (Gutentag and others, 1984) of the aquifer at a given location (higher K values correspond to larger radii). The radii were calculated using hypothetical ground-water flow and particle-tracking simulations. Four values for the radius were used: 300 meters, 1,300 meters, 3,000 meters, and 4,000 meters. These values were then assigned to each well depending on the hydraulic conductivity of the aquifer at that location. Each well was then assigned a bearing of the upgradient direction. The orientation was determined by subtracting 180 degrees from the regional ground-water flow direction at the well location. The ground-water flow direction was determined using the GRID function ASPECT (eight possible directions) on a raster data set of the water-table surface. An Arc Macro Language (AML) program was used to create a 90-degree sector for each well with the appropriate orientation and radius and then to summarize the amount of irrigated land within the sector using the GRID function FOCALSUM with the WEDGE option ([focalsum of irrigated cells/number of possible cells in wedge] x 100). After the amount of irrigated land was determined to be a statistically significant variable for the model, this amount was calculated for every location in the High Plains aquifer. A vector data set of polygons representing ranges in hydraulic conductivity in the aquifer (Cederstrand and Becker, 1998) was converted to a raster data set by using the above radii values to represent the various K ranges. The raster data set of the radii values was combined, using the GRID function COMBINE, with the aspect data set (calculated from the water-table surface) to determine the size and orientation of the 90-degree sector for every grid cell of an 80-meter grid. Once the wedge-shaped buffer size and orientation were determined, the buffer was used to summarize the amount of irrigated land within the buffer by using the GRID function FOCALSUM with the WEDGE option. This was calculated at each cell location creating and interpolated surface. To increase processing speed, a grid of the amount of irrigated land within a 90-degree sector was produced for each possible radius and orientation, resulting in 32 separate grids (4 radius values and 8 possible upgradient directions). The GRID function PICK was then used to choose the appropriate irrigated amount value from the possible 32 grids, depending on the radius value and flow direction at each cell. U.S. Geological Survey OFR 2002-441 20040401 The second variable used to calculate the probability map was the amount of nonirrigated land upgradient from the well location. The location of nonirrigated land was determined using the raster data set of location of irrigated lands (Qi and others, 2002) and the raster data set of the location of agricultural land in the High Plains area (U.S. Geological Survey, 1992). Because the location of agricultural land was used to limit where irrigated land was classified (a mask), the subtraction of the two raster data sets resulted in the classification of agricultural land that was not irrigated (nonirrigated land). This raster data set was then summarized using the 90-degree sector technique described above. The percentage of nonirrigated land was calculated within the 90-degree sector upgradient from each cell in an 80-meter raster data set. None 20040501 The third variable used to calculate the probability map was the amount of organic matter in soils of the High Plains study area. A vector polygon data set was used to represent the distribution of organic soil matter based on STATSGO soil-mapping units (U.S. Department of Agriculture, 1991) in the High Plains. For each mapping unit, the area-weighted values for various parameters, such as organic matter, are summarized depending on how much of the mapping unit is covered by a given soil type. Because the mapping units were large compared to the size of the 90-degree sectors, data were not summarized in this manner. The vector data set was converted to an 80-meter raster data set using the amount of organic matter in soil as the value of the cells. STATSGO 20040601 The fourth variable required to calculate the probability map was the depth to water. The depth to water was calculated using the measured depth to water at 8,514 well locations throughout the High Plains in the year 2000. The depth to water table raster data set was at a 500-meter resolution so the 90-degree sector method was not used because only a few cells would be within the largest wedge buffer used. The depth to water table raster data set was resampled to 80-meter resolution for use in the calculation of the probability map to match the resolution of the other input raster data sets. None 20040601 The final variable required to calculate the probability map was the percentage of clay in the unsaturated zone. The data set was created using approximately 56,000 lithologic logs from wells across the High Plains that were drilled deeper than the 2000 water table. The lithologic intervals described in the logs were compiled, simplified, and assigned terms of gravel, sand, clay, silt, or rock, or a combination of lithologies based on available lithologic descriptions from the logs. The thickness of each lithologic interval was calculated as a percentage of the total unsaturated-zone thickness. Kriging was used to geostatistically interpolate these percentages to create surfaces of the percentage of each lithology in the unsaturated zone of the study area. The surface representing the percentage of clay in the unsaturated zone was then used in the calculation of the probability map. None 20040701 Evaluation and selection of two final multivariate models were based on statistical significance, model fit, and predictive ability (Gurdak and Qi, 2006). The two final models represent the Northern High Plains (NHP model) and the combined Central and Southern High Plains (CHP and SHP model). The five variables discussed above were found to be the most significant and produced the best model fits for predicting the probability of detecting nitrate concentrations exceeding 4 mg/L (as N) in recently recharged ground water. The models produced probability equations of weighted variables and a constant, which were used to calculate the probability for nitrate exceedance. The equation for the NHP was Probability (NHP) = [e^-0.374+(0.023xNonirrigated land)+(0.017x Irrigated land)+(-1.487xSoil organic matter)] / [1 + e^-0.374+(0.023xNonirrigated land)+(0.017xIrrigated land)+(-1.487xSoil organic matter)]. The equation for the combined CHP and SHP was Probability (CHP and SHP) = [e^1.158+(-0.010xdepth to water)+(0.043xNonirrigated land)+(0.011xIrrigated land)+(-0.019xUnsaturated zone clay)] / [ 1 + e^1.158+(-0.010xdepth to water)+(0.043xNonirrigated land)+(0.011xIrrigated land)+(-0.019xUnsaturated zone clay)]. Map algebra in GRID was used to calculate the probability surfaces for the two models. The two surfaces were then merged to create a single probability map for the High Plains area. Because the probability values ranged from 0 to 1, the resulting surface was multiplied by 100 for the final raster data set. 20040801 Although the processing for creating this data set was done at 80 meters so as to balance the processing capabilities of the computers and the degradation our higher resolution data sets (30-m resolution), the data set was resampled to 500 meters. One of the input data sets (depth to water) used to calculate this data set, was at a 500-meter resolution, therefore, this data set should not be released at a higher resolution than 500 meters. 20060101 Raster Grid Cell 2647 1563 1 Albers Conical Equal Area 29.500000 45.500000 -96.000000 23.000000 0.000000 0.000000 row and column 500.000000 500.000000 meters North American Datum of 1983 Geodetic Reference System 80 6378137.000000 298.257222 The values of the raster dataset represent the percentage probability of nonpoint-source nitrate contamination (greater than 4 mg/L as N) of recently recharged (defined as less than 50 years) ground water in the High Plains aquifer. Gurdak, J.J., and Qi, S.L., 2006, Vulnerability of recently recharged ground water in the High Plains regional aquifer to nitrate contamination: U.S. Geological Survey Scientific Investigations Report 2006-5050, 39p. U.S. Geological Survey Ask USGS -- Water Webserver Team mailing address

507 National Center

Reston VA 20192 USA 1-888-275-8747 (1-888-ASK-USGS) http://answers.usgs.gov/cgi-bin/gsanswers?pemail=h2oteam&subject=GIS+Dataset+ds192_hp_npctprob Downloadable Data Although these data have been used by the U.S. Geological Survey, U.S. Department of the Interior, no warranty expressed or implied is made by the U.S. Geological Survey as to the accuracy of the data. The act of distribution shall not constitute any such warranty, and no responsibility is assumed by the U.S. Geological Survey in the use of these data, software, or related materials. The use of firm, trade, or brand names in this report is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey. The names mentioned in this document may be trademarks or registered trademarks of their respective trademark owners. ASCII spatial and attribute information Winzip 15.869 http://water.usgs.gov/GIS/dsdl/ds192_hp_npctprob.asc.zip ARCE spatial and attribute information Winzip http://water.usgs.gov/GIS/dsdl/ds192_hp_npctprob.e00.zip None: This dataset is provided by USGS as a public service. 20060502 U.S. Geological Survey Ask USGS -- Water Webserver Team mailing address

507 National Center

Reston VA 20192 USA 1-888-275-8747 (1-888-ASK-USGS) http://answers.usgs.gov/cgi-bin/gsanswers?pemail=h2oteam&subject=GIS+Dataset+ds192_hp_npctprob FGDC Content Standards for Digital Geospatial Metadata FGDC-STD-001-1998 local time None http://www.esri.com/metadata/esriprof80.html ESRI Metadata Profile http://www.esri.com/metadata/esriprof80.html ESRI Metadata Profile http://www.esri.com/metadata/esriprof80.html ESRI Metadata Profile