Get reliable science

Using very high resolution remote sensing to monitor invasive weeds is an increasingly high priority for rangeland researchers across the United States. In the western United States, invasive weeds such as medusahead (Taeniatherum caput-medusae (L.) Nevski) are transforming vegetation communities by establishing small localized infestations. Monitoring through detection and mapping of infestation is very important to understand the rate of infestation and spread. Recent developments in digital imaging, GPS technology, and computer technology have provided new opportunities to speed the collection, processing and storage of field data. This study focuses on how the rate of infestation across extensive rangelands can be accurately estimated using data collected within a strict protocol from GPS-positioned, digital color aerial and ground level photography. Images were collected using Nikkon 200, 16 megapixel fitted with a 24mm lens. This study employs a continuously recording GPS device, digital camera, and a computer to acquire and manage information. Photographs were taken at 5-second intervals while the aircraft flew a grid over the study area approximately 1500m above ground level. For ground level imagery, photographs were taken on transects where infestations were identified. The position of the aircraft was monitored at one-second intervals via two recording GPS units. Aerial and ground level images had a ground pixel resolution of 0.35 m and 0.0004717 m respectively. Ground photo locations were also used as training sites to assign spectral values (red, green and blue) for use in image classification. Images were classified using VegMeasure 2 software and analysis was supported by ArcGIS and GlobalMapper