2012 ENVI User Group: Public Data, Remote Sensing, & Invasive Plant Species
At the 2012 ENVI User Group, Jim Alford, California Department of Fish and Game, presented “Conversion of a Heads Up Digitizer”. This presentation focused on the use of public domain remotely sensed data to predict Taeniatherum caput-medusae (medusahead) infestations, and is a case study from the central California foothills. In Mr. Alford’s study region, Taeniatherum caput-medusae growth is cause for concern, since it aggressively overtakes other local grasses causing problems for grazing animals and wildfire mitigation.
Because wild lands managers are faced with increasingly constrained resources, Mr. Alford and his team have used this project to demonstrate the advantages and results of using remotely sensed data in identifying invasive plant infestations, in comparison to traditional field methods. For this study, Mr. Alford used free imagery from the National Agricultural Imagery Program (NAIP), under the United States Department of Agriculture, which acquires aerial imagery during the agricultural growing seasons in the continental U.S.
Mr. Alford and his team used the landcover classification tools of ENVI to determine areas of invasive grasses in New Hogan Lake, Calaveras County, California. This location is managed by the Army Corps of Engineers for recreation and flood control. There is an aggressive weed removal program already in place at the study site, but unfortunately the program is hampered by a lack of botanical staff due to budgetary constraints. To quickly identify areas where invasive grasses have taken control of the local flora, the team used the Normalized Difference Vegetation Index (NDVI) in ENVI to analyze the imagery. The NDVI algorithm analyzes multispectral data to understand vegetation distribution in an image. Following the image analysis and identification of invasive plant location, the team mapped the results using ArcGIS. With this information, the team was able to conclude that woody plants could be indentified to species and predict the presence of invasive grasses. Additionally, the results supported the thought that the invasive grasses are limited to soils with high late spring moisture.
The results of this study were use to assist in the mitigation of invasive grasses that can have damaging effects of flood plans and recreation areas. The methods and results of this study demonstrate that there are easy to use image analysis tasks that can have a positive impact on land management.
What other types of image analysis do you see impacting land management?