Name: Predicted Background Conductivity
Display Field: StreamName
Type: Feature Layer
Geometry Type: esriGeometryPolyline
Is View: true
Is Updatable View: true
Source Schema Changes Allowed: true
Sources
Description: <p style="margin-top:0px; margin-bottom:0px; font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"><span style="font-family:inherit; font-size:large;">The information displayed is based on </span><span style="font-family:inherit; font-size:large;">specific conductivity </span><span style="font-family:inherit; font-size:large;">predictions </span><span style="font-family:inherit; font-size:large;">for stream segments in the contiguous United States </span><span style="font-family:inherit; font-size:large;">from the Predicted Background Conductivity (PBC) model. The PBC model was developed using a random forest modeling approach and enables comparison with measured in-stream conductivity. Geology, soil, vegetation, climate and other empirically measured data were used as inputs. The PBC model was designed for streams with natural background SC < 2000 µS/cm. Above this level (typical for freshwater), PBC model estimates may be less reliable. Data for some parameters that affect background SC were not readily available and were therefore not included in the model. These include freshwater and marine interfaces, natural mineral springs, salt deposits which may affect groundwater and streams, and other natural sources of salts. In such areas, the model is likely to underestimate background SC, but in some arid areas the model may need to be calibrated by as much as 300 </span><span style="font-family:inherit; font-size:large;">µS/cm (See Model validation below)</span><span style="font-family:inherit; font-size:large;">.</span><font size="3" style="font-family:inherit;"> </font><span style="font-family:inherit; font-size:large;">Local knowledge is often necessary to assess differences between predicted and measured background SC. </span><span style="font-family:inherit; font-size:large;">More information about the model and datasets can be found at </span><a href="https://epa.maps.arcgis.com/home/item.html?id=13fe4698f35d4a95ba9dfe6f8033f735" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit; font-size:large;">Freshwater Explorer Metadata</a><span style="font-family:inherit; font-size:large;">.</span></p><p style="margin-top:0px; margin-bottom:0px; font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"><span style="font-family:inherit; font-size:large;"> </span></p><p style="margin-top:0px; margin-bottom:0px; font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"><span style="font-family:inherit; font-size:large;"><strong>General Modeling Approach</strong></span></p><p style="margin-top:0px; margin-bottom:0px; font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"><span style="font-family:inherit; font-size:large;">The <a href="https://github.com/USEPA/StreamCat" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">StreamCat dataset</a> and process was used to develop stream-specific model predictions (Hill et al. 2016) based on watershed averages for </span><span style="font-family:inherit; font-size:large;">each NHD+ segment in the contiguous United States </span><span style="font-family:inherit; font-size:large;">(McKay et al. 2012). These stream segments drain an average area of 3.1 square kilometers which characterizes the spatial grain size of this dataset. Natural background SC was not estimated for streams shown as grey line<span style="font-family:inherit;">s. </span></span><span style="font-family:inherit; font-size:large;">The empirical background conductivity model was developed using the following steps:</span></p><p style="margin-top:0px; margin-bottom:0px; font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"> </p><ol><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">Create training and validation data sets of SC </span><span style="font-family:inherit; font-size:large;">observations from minimally altered stream segments.</span></span></li><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">Characterize temporally and spatially specific watershed </span><span style="font-family:inherit; font-size:large;">environments for each observation, including antecedent</span></span></li><li><span style="font-family:inherit; font-size:large;">conditions.</span></li><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">Relate observed SC to environmental predictors using a </span><span style="font-family:inherit; font-size:large;">machine learning technique (random forests).</span></span></li><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">Assess model performance and validate using multiple </span><span style="font-family:inherit; font-size:large;">observations made at randomly chosen stream segments.</span></span></li></ol><p style="margin-top:0px; margin-bottom:1.5rem; font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"><span style="font-family:inherit; font-size:large;">For a detailed description of the step used to develop the PBC model, see <em>Olson and Cormier, 2019.</em></span></p><div style="font-family:"Avenir Next W01", "Avenir Next W00", "Avenir Next", Avenir, "Helvetica Neue", sans-serif; font-size:16px;"><span style="font-family:inherit; font-size:large;"><em><br /></em></span><p style="margin-top:0px; margin-bottom:1.5rem;"> </p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:medium;"><strong>Training and Validation Data</strong></span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">The t</span><span style="font-family:inherit; font-size:large;">ra</span><span style="font-family:inherit; font-size:large;">ining and validation datasets </span><span style="font-family:inherit; font-size:large;">consist of minimally disturbed sites. More than 2.4 million SC observations were obtained from the Water Quality Portal (WQP) (USEPA 2016b), state natural resource agencies, the U.S. Geological Survey (USGS) National Water Information System (NWIS) system (USGS 2016), and data used in Olson and Hawkins (2012). Data were downloaded from the </span><a href="https://www.waterqualitydata.us/" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank"><span style="font-family:inherit; font-size:large;">WQP website</span></a><span style="font-family:inherit; font-size:large;"> using the following query criteria: </span></span></p><p style="margin-top:0px; margin-bottom:0px;"> </p><ul><li><span style="font-family:inherit; font-size:large;">Country - United States.</span></li><li><span style="font-family:inherit; font-size:large;">Sample Media - Water.</span></li><li><span style="font-family:inherit; font-size:large;">Characteristics - Conductivity, Specific Conductivity, Specific Conductance, Calculated/Measured Ratio. </span></li><li><span style="font-family:inherit; font-size:large;">Date range - observations between 1 January 2001 and 31 December 2015. This time period was chosen so that <a href="https://modis.gsfc.nasa.gov/data/" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data</a> (NASA 2019) could be used as predictors in the model. </span></li></ul><p style="margin-top:0px; margin-bottom:1.5rem;"> </p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;">During development, 56 potential explanatory parameters were evaluated. The original source data and final datasets are available for download <a href="https://doi.org/10.23719/1500945" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">here</a>. Predicted Background Conductivity metadata and data are also available on the Geoplatform.</span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;"> </span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">Each observation was related to the nearest stream segment in the NHD+. Data were limited to one observation per stream segment per month. SC observations with ambiguous locations and repeat measurements along a stream segment in the same month were discarded. Using estimates of anthropocentric stress derived from the StreamCat database (Hill et al. 2016), segments were selected with minimal amounts of human activity (Stoddard et al. 2006) using criteria developed for each Level II Ecoregion (Omernik and Griffith 2014). Stream segments were considered as minimally stressed when the associated watershed drainage area had </span><em>≤</em><span style="font-family:inherit; font-size:large;">0.5% impervious surface, </span><em>≤</em><span style="font-family:inherit; font-size:large;">5% urban, </span><em>≤</em><span style="font-family:inherit; font-size:large;">10% agriculture, and population densities of 0.8 to 30 people per square kilometer. Watersheds displaying large residuals during initial model predictions were assessed for evidence of other human activities not represented in StreamCat (e.g., mining, logging, grazing, or oil/gas extraction). Disturbed watersheds with a tidal influence or unusual geologic conditions, such as hot springs, were not removed from the dataset. Some sites with high levels remain in the dataset (e.g., mining influenced but no easily accessible evidence). Some sites with high SC remain in the dataset (e.g., mining influenced but no national record). About 5% of SC observations in each National Rivers and Stream Assessment (NRSA) region were then randomly selected as independent validation data. The remaining observational dataset was used for model calibration.</span></span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit;"> </span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;">The final training data set used for modeling had 1785 stream segments with 11 796 observations, and the validation data set had 92 segments with 581 observations. The majority of segments had a single observation but ranged up to 165 observations per segment.</span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;"> </span></p><p style="margin-top:0px; margin-bottom:0px;"><strong><span style="font-family:inherit; font-size:medium;">Model Validation</span></strong></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;">The "Model Validation" view on the Freshwater Explorer shows the predictive performance at reference sites that were used to develop PBC model. In the wetter and more forested areas, reference sites were more abundant and predictions are more precise. In the central U.S., fewer reference sites were available and predictions are more uncertain. In the grass and shrub lands east of the continental divide, measured SC was over-predicted by the model by more than 100 µS/cm at 5% of sites (yellow dots) and under-predicted by more than 100 µS/cm at 3% of sites (red dots). Calculated differences between measured and predicted SC are reported as residuals in pop-up boxes on this view. There are many potential causes for these differences, including data reporting errors and reference site reliability. </span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;"> </span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit; font-size:large;">Overall, the model explained most of the variation in SC and produced reasonably accurate predictions for both calibration data (Mean Absolute Error = 22 µS/cm, Nash-Sutcliffe Efficiency = 0.92, and Coefficient of Determination = 0.92) and external validation data (Mean Absolute Error = 29 µS/cm, Nash-Sutcliffe Efficiency = 0.87, and Coefficient of Determination = 0.87). Values reported as background only apply to streams and have not been validated for lakes or wetlands.</span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit;"> </span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">More details can be found in <em>Olson and Cormier (2019)</em>.</span><br /></span></p><p style="margin-top:0px; margin-bottom:0px;"><span style="font-family:inherit;"> </span></p><p style="margin-top:0px; margin-bottom:0.0001pt;"><strong><span style="font-family:inherit; font-size:large;">References</span></strong></p><ul><li><span style="font-family:inherit; font-size:large;">Hill, R.A., Weber, M.H., Leibowitz, M.H., Olsen, A.R., Thornbrugh, D.J., 2016. The Stream-Catchment (StreamCat) Dataset: A Database of Watershed Metrics for the Conterminous United States. JAWRA Journal of the American Water Resources Association, 52(1), 120-128. doi: <a href="https://doi.org/10.1111/1752-1688.12372" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">https://doi.org/10.1111/1752-1688.12372</a></span></li><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">McKay, L., Bondelid, T., Dewald, T., Johnston, J., Moore, R., Rea, A., 2012. NHDPlus Version 2: User Guide. National Operational Hydrologic Remote Sensing Center, Washington, DC. Available at: </span><a href="https://epa.maps.arcgis.com/home/item.html?id=540abb1d015b4bd2b87d30f4c28a58cb&view=table" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank"><span style="font-family:inherit; font-size:large;">https://nctc.fws.gov/courses/references/tutorials/geospatial/CSP7306/Readings/NHDPlusV2_User_Guide.pdf</span></a></span></li><li><span style="font-family:inherit; font-size:large;">NASA, 2019. Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data. Available at: <a href="https://epa.maps.arcgis.com/home/item.html?id=540abb1d015b4bd2b87d30f4c28a58cb&view=table" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">https://modis.gsfc.nasa.gov/data/</a></span></li><li><span style="font-family:inherit; font-size:large;">Olson, J.R., Hawkins, C.P., 2012. Predicting natural base‐flow stream water chemistry in the western United States. Water Resources Research, 48(2). doi: <a href="https://epa.maps.arcgis.com/home/item.html?id=540abb1d015b4bd2b87d30f4c28a58cb&view=table" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">https://doi.org/10.1029/2011WR011088</a></span></li><li><span style="font-family:inherit; font-size:large;">Olson, J.R. and Cormier, S.M., 2019. Modeling spatial and temporal variation in natural background specific conductivity. Environmental Science and Technology. doi: <a href="https://epa.maps.arcgis.com/home/item.html?id=540abb1d015b4bd2b87d30f4c28a58cb&view=table" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">https://dx.doi.org/10.1021/acs.est.8b0677</a>7</span></li><li><span style="font-family:inherit; font-size:large;">Olson, J.R. and Cormier, S.M., 2019. Modeling spatial and temporal variation in natural background specific conductivity: Data sets and R-code. doi: <a href="https://doi.org/10.23719/1500945" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">https://doi.org/010.23719/1500945</a></span></li><li><span style="font-family:inherit; font-size:large;">Omernik, J. M.; Griffith, G. E., 2014. Ecoregions of the conterminous United States: evolution of a hierarchical spatial framework. Environmental. Management. 54 (6), 1249−1266. doi: <a style="color:rgb(0, 121, 193); font-family:inherit;" target="_blank">https://doi.org/10.1007/s00267-014-0364-1</a></span></li><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">U.S. Environmental Protection Agency (U.S. EPA), 2016. STORET. Available at: </span><a href="https://epa.maps.arcgis.com/home/item.html?id=540abb1d015b4bd2b87d30f4c28a58cb&view=table" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank"><span style="font-family:inherit; font-size:large;">https://www.epa.gov/storet/</span></a></span></li><li><span style="font-family:inherit; font-size:large;">U.S. Geological Survey (USGS), 2016. National Water Information System. Available at: <a href="https://epa.maps.arcgis.com/home/item.html?id=540abb1d015b4bd2b87d30f4c28a58cb&view=table" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank">https://waterdata.usgs.gov/nwis</a></span></li><li><span style="font-family:inherit;"><span style="font-family:inherit; font-size:large;">Stoddard, J. L., Larsen, D. P., Hawkins, C. P., Johnson, R. K., Norris, R. H., 2006. Setting expectations for the ecological condition of streams: the concept of reference condition. Ecol. Appl. 2006, 16 (4), 1267−1276. doi: </span><a href="https://doi.org/10.1890/1051-0761(2006)016[1267:SEFTEC]2.0.CO;2" style="color:rgb(0, 121, 193); text-decoration-line:none; font-family:inherit;" target="_blank"><span style="font-family:inherit; font-size:large;">https://doi.org/10.1890/1051-0761(2006)016[1267:SEFTEC]2.0.CO;2</span></a></span></li></ul></div>
Copyright Text: John O. Olson, Christopher L. Wharton, Susan M. Cormier
Min. Scale: 1879005
Max. Scale: 0
Default Visibility: true
Max Record Count: 1000
Supported query Formats: JSON, geoJSON, PBF
Use Standardized Queries: True
Extent:
XMin: -2909390.27397196
YMin: 167800.040432617
XMax: 2937574.67842609
YMax: 3254036.20985305
Spatial Reference: 102965 (6350)
Drawing Info:
{"renderer":{"authoringInfo":{"type":"classedColor","classificationMethod":"esriClassifyManual"},"type":"classBreaks","field":"AvgPredEC","legendOptions":{"order":"ascendingValues"},"minValue":10.293370165745854,"classBreakInfos":[{"symbol":{"color":[115,0,76,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":25,"label":"0 - 25"},{"symbol":{"color":[112,35,90,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":50,"label":"25 - 50"},{"symbol":{"color":[110,56,107,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":75,"label":"50 - 75"},{"symbol":{"color":[104,75,125,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":100,"label":"75 - 100"},{"symbol":{"color":[94,93,143,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":125,"label":"100 - 125"},{"symbol":{"color":[82,113,163,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":150,"label":"125 - 150"},{"symbol":{"color":[61,132,184,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":200,"label":"150 - 200"},{"symbol":{"color":[55,149,196,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":250,"label":"200 - 250"},{"symbol":{"color":[96,163,181,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":300,"label":"250 - 300"},{"symbol":{"color":[130,179,169,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":350,"label":"300 - 350"},{"symbol":{"color":[160,194,155,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":400,"label":"350 - 400"},{"symbol":{"color":[187,209,138,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":450,"label":"400 - 450"},{"symbol":{"color":[215,227,125,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":500,"label":"450 - 500"},{"symbol":{"color":[243,245,108,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":550,"label":"500 - 550"},{"symbol":{"color":[252,231,91,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":600,"label":"550 - 600"},{"symbol":{"color":[252,196,76,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":650,"label":"600 - 650"},{"symbol":{"color":[252,164,63,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":700,"label":"650 - 700"},{"symbol":{"color":[250,133,50,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":800,"label":"700 - 800"},{"symbol":{"color":[245,99,37,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":900,"label":"800 - 900"},{"symbol":{"color":[240,67,29,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":1000,"label":"900 - 1000"},{"symbol":{"color":[232,16,20,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":1500,"label":"1000- 1500"},{"symbol":{"color":[173,16,20,255],"width":1.7499999999999998,"type":"esriSLS","style":"esriSLSSolid"},"classMaxValue":5000,"label":"1500 - 5000"}],"classificationMethod":"esriClassifyManual"},"transparency":0}
HasZ: true
HasM: true
Has Attachments: false
Has Geometry Properties: true
HTML Popup Type: esriServerHTMLPopupTypeAsHTMLText
Object ID Field: OBJECTID
Unique ID Field:
IsSystemMaintained : True
Global ID Field: GlobalID
Type ID Field:
Fields:
- OBJECTID (type: esriFieldTypeOID, alias: OBJECTID, SQL Type: sqlTypeOther, length: 0, nullable: false, editable: false)
- COMID (type: esriFieldTypeInteger, alias: COMID, SQL Type: sqlTypeOther, nullable: true, editable: true)
- StreamName (type: esriFieldTypeString, alias: StreamName, SQL Type: sqlTypeOther, length: 65, nullable: true, editable: true)
- AvgPredEC (type: esriFieldTypeDouble, alias: Conductivity (µS/cm), SQL Type: sqlTypeOther, nullable: true, editable: true)
- F50Pct_PL (type: esriFieldTypeDouble, alias: Conductivity (µS/cm), SQL Type: sqlTypeOther, nullable: true, editable: true)
- FTYPE (type: esriFieldTypeString, alias: Stream Type, SQL Type: sqlTypeOther, length: 24, nullable: true, editable: true)
- FCODE (type: esriFieldTypeInteger, alias: FCODE, SQL Type: sqlTypeOther, nullable: true, editable: true)
- GlobalID (type: esriFieldTypeGlobalID, alias: GlobalID, SQL Type: sqlTypeOther, length: 38, nullable: false, editable: false)
- Shape__Length (type: esriFieldTypeDouble, alias: Shape__Length, SQL Type: sqlTypeDouble, nullable: true, editable: false)
Templates:
Name: Predicted Background Conductivity
Description:
Drawing Tool: esriFeatureEditToolNone
Prototype:
Is Data Versioned: false
Has Contingent Values: false
Supports Rollback On Failure Parameter: true
Last Edit Date: 5/24/2023 3:24:46 PM
Schema Last Edit Date: 5/24/2023 3:24:46 PM
Data Last Edit Date: 7/27/2021 1:55:52 PM
Supported Operations:
Query
Query Top Features
Query Analytic
Query Bins
Generate Renderer
Validate SQL
Get Estimates
ConvertFormat