USING QUICKBIRD SATELLITE IMAGERY TO DISTINGUISH TWO AQUATIC WEEDS IN SOUTH TEXAS

Authors: Everitt, James; Yang, Chenghai

Submitted to: Journal of Aquatic Plant Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/10/2006
Publication Date: 4/1/2007
Citation: Everitt, J.H., Yang, C. 2007. Using QuickBird satellite imagery to distinguish two aquatic weeds in south Texas. Journal of Aquatic Plant Management. 45:25-31.

Interpretive Summary: Invasive weeds are an extremely big problem in the United States and cost its citizens approximately $35 billion annually. Nowhere are these biological invasions more evident than in rivers, lakes, and reservoirs. Waterhyacinth and waterlettuce are two exotic, invasive aquatic weeds that invade and clog waterways in the southern half of the United States. Research was conducted to determine the potential of using QuickBird satellite imagery for distinguishing waterhycinth and waterlettuce infestations in a large reservoir in south Texas. Three subsets of the satellite image were extracted and used as study sites. Waterhyacinth occurred in all three subset images, whereas waterlettuce was in only one subset image. Accuracy assessments performed on supervised and unsupervised classification maps of images from the three sites had producer's and user's accuracies for waterhyacinth ranging from 73% to 100%. Accuracy assessments performed on supervised and unsupervised classification maps from an image of one site had producer's and user's accuracies for waterlettuce ranging from 82% to 100%. These results should be of interest to wetland resource managers.

Technical Abstract: QuickBird false color satellite imagery was evaluated for distinguishing waterhyacinth [Eichhornia crassipes (Mort.) Solms] and waterlettuce (Pistia stratiotes L.) infestations in a large reservoir in south Texas. The imagery had three bands (green, red, and near-infrared) and contained 11-bit data. Three subsets of the satellite image were extracted and used as study sites. Waterhyacinth occurred in all three subset images, whereas waterlettuce was in only one subset image. Supervised and unsupervised classification techniques were used to classify the imagery. Accuracy assessments performed on supervised classification maps of images of the three sites had producer's and user's accuracies for waterhyacinth ranging from 73% to 100%, while accuracy assessments performed on unsupervised classification maps of images of the three sites had producer’s and user’s accuracies for waterhyacinth ranging from 74% to 100%. An accuracy assessment performed on a supervised classification map of an image from only one site showed that waterlettuce had both a producer's and user's accuracy of 100%, while an accuracy assessment performed on a unsupervised classification map of an image from the same site showed that waterlettuce had a producer's and user's accuracy of 82% and 90%, respectively. These results indicate QuickBird satellite imagery coupled with image analysis techniques can be used successfully for detecting waterhyacinth and waterlettuce infestations.