Historic Rice Fields of South Carolina South Carolina Rice Kingdom Historic Rice Fields of South Carolina 2018-08-01 https://services1.arcgis.com/x5wCko8UnSi4h0CB/arcgis/rest/services/Rice_fields/FeatureServer R. Daniel Hanks Clemson University Palmetto Green Fellow rhanks@clemson.edu Clemson University, Department of Forestry and Environmental Conservation, 261 Lehotsky Hall Clemson South Carolina 29634 US (864) 656-0681 We sought to identify and digitize historic rice fields in seven coastal counties of South Carolina. Using ArcGIS Online (AGOL) we developed an application to use as a workspace for headsup digitizing of rice fields. The application allowed us to identify rice fields using a combination of remotely sensed data sources that included satellite imagery (ArcGIS Online World Imagery MapServer), aerial photographs, National Wetlands Inventory (Wilen and Bates, 1995), USGS topographic maps, and LiDAR generated digital elevation models (DEMs) and intensity. Rice fields were identified by the visual presence in remotely sensed data of channels, dykes, and quarter drains, which are the main features associated with rice fields. The linear nature of these features associated with rice fields is in sharp contrast to the more typical tree-like branching associated with natural drainage patterns and were the major factor used to identify landscape features as rice fields. Once a rice field was identified we manually digitized the field using the AGOL application. During heads up digitizing, using the AGOL application, a mapper’s confidence in whether an area identified as a rice field was or was not a rice field by scoring the field as a 1 (low confidence), 2 (moderate confidence), or 3 (high confidence). The identification, digitization, and confidence quantification processes were performed at least twice by two different mappers across the entire coastal plains of South Carolina. Once the digitization process was complete we developed a procedure to field validate areas identified as rice fields. It was impossible, in a reasonable time frame, to field validate each of the digitized rice fields due to the extensive geographic area that the rice fields cover. So, to ensure a statistically robust method of selecting a representative subset of rice fields to field validate we developed a concept model and subsequent protocol that is systematic and reproducible. This methodology allowed for a randomly selected subset of rice fields for in person visits and field validation. Accessibility was a major concern for field validation due to private ownership of rice fields or fields being located great distances from roads or navigable waters. Therefore, in order to be included in our subset of rice fields for validation, we initially filtered rice fields based on accessibility. Rice fields were considered accessible if they were within 200 m of a public road or navigable waters. This step ensured ease of access and reduced the amount of time required to validate a rice field thereby allowing for more fields to be assessed, resulting in improved spatial distribution and increased coverage. We then randomly selected a percentage of the rice fields that met the accessibility requirement within each confidence category (i.e., high, medium, low). We elected to field validate a greater number of rice fields we identified but with low confidence and fewer of those we were more confident were rice fields during the digitization and classification process. From the subset of accessible rice fields we randomly selected 325 (10.8 % of all identified fields) to field validate. Of 325 randomly selected fields, 20.3 % were classified as low confidence, 12.6 % were classified as medium confidence, and 9.9 % were classified as high confidence. Field validation consisted of in person visitation to each of the randomly selected rice fields. We field validated each of the selected rice fields. Field validation was accomplished by in person visits to the randomly selected sites where identification and confirmation of dikes and other key identifying features were assessed. Fields were identified as either being a rice field or not (i.e., Yes or no). Fields that were identified as a rice field (i.e., Yes) were included in our final map and summary statistics of the rice fields. Additionally, photographs were taken of at least one of the identifying features at each site. These images were added to the AGOL application for further confirmation (if in question). R. Daniel Hanks, Richard Coen, Robert F. Baldwin, Michael Gouin, Travis Folk, and Ernie Wiggers Rice fieldsmapsSouth CarolinaHistoric Rice FieldsSouth Carolina Rice KingdomWetlands Wetland landscapes land use transformation slavery ecosystem pathways land use legacies South Carolina United States Southeastern United States -83.6192 -78.2359 31.7618 35.3317 R. Daniel Hanks Clemson University Palmetto Green Fellow rhanks@clemson.edu Clemson University, Department of Forestry and Environmental Conservation, 261 Lehotsky Hall Clemson South Carolina 29634 US https://services1.arcgis.com/x5wCko8UnSi4h0CB/arcgis/rest/services/Rice_fields/FeatureServer Historic Rice Fields of South Carolina Historic Rice Fields of South Carolina 2018-08-01 Historic Rice Fields of South Carolina 2018-08-01 FGDC CSDGM Metadata 2019-04-04 11:02:46.37 2019-04-05 08:48:59.39 1.0 FGDC editor:esri.dijit.metadata.editor 2019-04-04 aae163eb419c4b0a8147202c50ad67c0 https://services1.arcgis.com/x5wCko8UnSi4h0CB/arcgis/rest/services/Rice_fields/FeatureServer