Methods detail - GIS
1. Software platform: QGIS
We use the free software QGIS for most mapping. Google Maps and Google Earth have narrow roles in preparing data for use with QGIS.
We have three levels of mapped geographic data:
- Statewide simple, all statewide data except excludes NRCS soil maps.
- Three substate vertical strips, all data including about one third of upstate county soil maps, overlapping in counties with the adjacent strip(s).
- Categorical site. All data except omitting PSUR pesticide use, narrowed soils to immediate and adjacent counties, add LIDAR and property tax lot boundaries. There is one map file per cluster of sites.
The statewide simple QGIS set was used to produce maps for the categoricals-detail page in this set.
We use the substate vertical strips for long term site and lake evaluation. We do not prepare site map files for long term groundwater sites or lakes.
2. Statewide data
These data are all from public sources, often the New York State GIS clearinghouse.
| Layer(s) | Source | Description |
|---|---|---|
| Topography | USGS | Scanned topographic quad images. Loaded from an online source. Used for basemaps and to read elevations of nearby surface water bodies versus candidate well installation sites. |
| Land use | US EPA | MRL land use based on satellite data. |
| Agricultural land use | USDA | Cropscape. Used for vineyard, fruit, and vegetable field area overlaying on soils and surficial geology. |
| Elevation | Digital elevation maps 10m. | |
| Hydrography | NYSDEC? | |
| 14 digit HUC watersheds | ||
| CSLAP lakes | NYSDEC | This is a point map with descriptive info about the lake having a volunteer group. |
| Pesticide use and sales at zip code level | NYSDEC/Cornell | Tabular data joined with US Census ZCTA polygons to have geometry. Data are in SQLite3 database, consisting of annual data appended with a year number (no years in source records). Since these are itemized by year, sales or use, active ingredient, and zip code, there is potential for mapping individual active ingredients, sums of all AIs or groups of AIs, weighted sums. |
| County boundaries | US Census | |
| Digital aerial photographs | New York State | Color and color IR. Various dates. |
| Roads | US Census | |
| Active railways | NYS DOT |
3. Google Earth for locations of candidate categorical sites
We apply the Windows desktop application of Google Earth Pro.
This has access to Google’s web crawler data; the locations of sites are address-geocoded by Google when used in Google Earth. We searched by “greenhouses near Seneca County, NY”, for example, appending the results (as KML format) county by county until all counties of the Upstate zone were covered. The resulting KML file can be edited to remove duplication and obvious false positive results. For example, the word “nurseries” finds child care facilites as well as plant nurseries.
This provided good results for sod farms, golf courses, fruit and vegetable farms (using extra terms like apples, orchards), greenhouses, outdoor nurseries, and vineyards. It does not yield railway or power rights of way, of course. It can’t find “other turfgrass” sites which are too diverse for searching.
Since these maps are based on data the site owners publish on the public world wide web to promote themselves, these map layers are inherently not confidential.
The composite categorical candidate maps are the only copyrighted information in the GIS mapset. Google allows use in publications with credit to them, as we do in this website.
4. County by county data
NRCS “SSURGO” soils maps with descriptive tables, for each upstate county, are downloaded from the NRCS official soil map site.
We use them beyond their recommended scale by looking at soils on individual properties. However, there was only one case where the mapped soil was very different from what was on the ground. The soil map had been done before 2000. Some time between soil map time and 2000 when the current owner bought the property, nearly all of the site had its topsoil and subsoil to a few meters was “scalped” and removed. The new “topsoil” was a clay layer formerly several meters deep.
For some categorical groundwater sites, the site-level QGIS collection contains two more maps obtained from State or County GIS repositories:
- LIDAR images or or high resolution topography contours. This is used in refining depth to groundwater estimates while planning to locate wells.
- Property boundaries, with ownership.
These data are all non-confidential.
5. Site level map files
A site level QGIS map is simply a copy of one of the three vertical strip maps with most county soil maps stripped for faster loading, LIDAR and property maps for the county when needed, and saved to zoom to the exact property and its immediate adjacent properties when the map file is opened. These QGIS maps are confidential because of the zooming in to the individual property.
These are prepared only for categorical groundwater sites. We use the three larger strip maps for longterm groundwater sites and lakes, usually not saving their zoomed-in locations.
6. Cooperator and sampling point locations
Locations and identities of all groundwater cooperators are confidential from NYSDEC and the public, known only to Cornell SWL and the cooperator. Lakes and lake sampling points are not confidential.
When we have reached agreement with a categorical site cooperator, we digitize one point for their property plus one point for each sampling point. The digitizing of sampling points is done by locating the point on a zoomed-in digital aerial photo in QGIS using landmarks and memory of the site visit; points can also be marked on a printed digital aerial photo while walking the site with the owner. These points are accurate to within a few meters.
Since we have only one sampling point per long term cooperator, there is no distinction between a property point and a sampling point.
The digitizing of cooperator property locations, either long term or categorical, is typically from Google Maps web application which again uses a zoomed-in aerial photo. These are less accurate than sampling point locations since they are intended only for confidential upstate-wide maps for sampling route planning.
When we portray site locations on public maps such as within this website, we color in the county without indicating the site category. In tables showing categoricals individually, locations are listed as the ecological region. In tables of long term cooperators, we list the county and whether the property is a business, residence, or governmental.
For geographic locations in EQuIS database submissions (see tabular database), we probably will report the zip code centroids for long term sites as we did in earlier years, and we are considering options for the categorical sites. One possibility would be to report the centroid of the ecological region containing the categorical site. This will require registering a new location type in EQuIS. EQuIS also forms its sampling location identifications using county codes, which is acceptable blurring for our long term sites but inconsistent with anonymity of some categorical sites.
We hope that these approaches to location blurring will be comfortable for all groundwater site cooperators.
Lake sampling point locations are typically digitized by a lake volunteer for lakes where volunteers do the sampling. Cornell SWL sampling points in lakes or tributaries are done on zoomed-in digital aerial photos, in Google Maps or the online digital aerials accessed via QGIS.
The digitized sampling points can be used to prepare maps to include in reports to the confidential cooperator or lake volunteer.
7. Potential for GIS mapset use by others
With the cooperator and sampling point locations maps deleted, the GIS collection is based on public data and can be provided to NYSDEC staff for use via QGIS. (This is not a contract deliverable.) Typically Cornell has prepared interpretive maps on NYSDEC request rather than providing GIS files.
Excerpts from the non-confidential components of the mapset are also used in Cornell teaching, notably in an undergraduate groundwater course. In one semester we were able to have the students learn and use QGIS.
Last updated 2023-08-17, sp17 AT cornell.edu.