Identifying Land Use/Land Cover (LULC) Using National Agriculture Imagery Program (NAIP) Data as a Hydrologic Model Input for Local Flood Plain Management
MetadataShow full metadata
The purpose of this study was to explore the utility of remotely sensed data acquired for agricultural applications to assist urban planners in land use and land cover (LULC) classifications. The National Agriculture Imagery Program (NAIP) offers local planners a high resolution (i.e. one-meter), multispectral (4 bands: red, blue, green and near infrared) dataset at little (or no) cost. NAIP imagery was selected because of its low cost and potential for small scale land use and land cover classifications similar to the success Landsat (30 meter, multispectral: 4 band) imagery has achieved with large scale classifications. The study was conducted using a subdivision in South-central Texas (i.e. El Camino Real) and the surrounding (rural) property. Supervised (parametric and non-parametric) classification procedures were conducted on the El Camino Real subset using ERDAS Imagine 9.3®. Stratified random sample points were generated for accuracy assessment via a ground based visual assessment of each point's LULC class. By using a 7 class LULC schema, a supervised classification of the NAIP imagery resulted in classification accuracy of 86%. When the schema was reduced to two broad classes (i.e. impervious and pervious cover), the classification accuracy climbed to 95%. These results suggest the need for a continued exploration of NAIP data utility for local planning purposes.