The use of unsupervised classification of Landsat-5 TM images in analysing the types of vegetation in the areas of "breaking dikes or opening sluice for water store" / 中华流行病学杂志

<p><b>OBJECTIVE</b>To clarify the change of vegetation types and its relationship between the density of alive-snails in the areas of "breaking dikes or opening sluice for water store" in Jicheng.</p><p><b>METHODS</b>Synthesized false color images of Jicheng before and after 1998 (1994 and 2003) were classified without supervision and results were compared. Vegetation types were identified on the spot.</p><p><b>RESULTS</b>Normalized difference vegetation index (NDVI) of snail habitats before 1998 were between 126 and 183 in Jicheng, whose vegetation types were mainly paddy, cotton and cabbage. NDVI of snail habitats in Jicheng after 1998 were between 152 and 193 whose vegetation types were mainly poplar forest, bulrush and grass. Areas of snail habitats increased from 64.64% to 66.47%. Snail habitats were mostly composed of mixed vegetation types and mono-typed vegetation was hardly found. According to the density of alive-snails orders from high to low were poplar forest and bulrush, poplar forest and grass, bulrush.</p><p><b>CONCLUSION</b>Vegetation types would not be identified by unsupervised classification only. Poplar forest, bulrush and grass were closedly related to the density of alive-snails.</p>

Application of "time series analysis" in the prediction of schistosomiasis prevalence in areas of "breaking dikes or opening sluice for waterstore" in Dongting Lake areas, China / 中华流行病学杂志

<p><b>OBJECTIVE</b>To provide the fittest model for forecasting schistosomiasis prevalence in Haokou village of "breaking dikes or opening sluice for waterstore" in Dongting Lake areas by comparing the results of Moving Average, Exponential Smoothing, Autoregressive Model and Autoregressive integrated moving average model (ARIMA model) from 1990 to 2002.</p><p><b>METHODS</b>Error sum of square of four statistical methods was compared and the fittest model was chosen.</p><p><b>RESULTS</b>Error sum of square of predicted schistosomiasis prevalence rates in Haokou village from 1994 to 2002 were 39.40, 39.86, 26.63, 22.54 respectively.</p><p><b>CONCLUSION</b>ARIMA model seemed to be the fittest one in the prediction of schistosomiasis prevalence in Haokou village of "breaking dikes or opening sluice for waterstore" in Dongting Lake from 1990 to 2002.</p>