Remote sensing for monitoring on the health of artificial Robinia Pseudoacacia Forests in the Yellow River Delta

Some of Robinia Pseudoacacia Forests have been dead or dieback in the Yellow River Delta Since 1990s. We evaluated 376 trees at 40 sites across the study area using the USFS Crown Condition Rating Guide, and separated Robinia Pseudoacacia into four levels of tree vigor. The area of Robinia Pseudoacacia was extracted from Landsat TM image using supervised classification. NDVI, NDWI, SAVI, MSAVI and TCT transform were used, and then cluster analysis was used to separate the Robinia Pseudoacacia into four class of tree vigor. The field data were used to measure the accuracy of various vegetation index classification techniques. The Normalized Difference Water Index (NDWI) transform provided the best overall accuracy, 82.5%, for classifying Robinia Pseudoacacia according to tree vigor. Health detection result and relative environment factors analysis showed that soil salinity and groundwater depth were found to be significantly correlative with the health of Robinia Pseudoacacia.