A Flex and ArcGIS Server based system for farmland environmental quality assessment and prediction in an agricultural producing area

•A Flex and ArcGIS Server based farmland environmental quality assessment and prediction system was developed.•Single factor index and weighted pollutant index method were used to establish environmental quality assessment model.•The average weighted pollutant index was 0.54, indicating good soil environmental quality in the research area.•An exponential smoothing forecasting model was established to predict heavy metal pollution.•The accuracy of the prediction models was about 90%. A farmland environmental quality assessment and prediction system was developed based on the database technology, WebGIS and model theory to provide intuitional information of farmland environmental pollution situation and the pollution trend for scientific decision making regarding monitoring soil quality. The designed system could be used for the query of agricultural environmental information, assessment of farmland environmental quality, and forecasting of farmland environmental pollution. An environmental quality assessment model was developed and integrated in the system by single factor index and weighted pollutant index method. The application of the assessment model showed that the single factor index of soil heavy metal Hg in the study area reached light pollution levels. The pollutant area occupied 21.3% of the entire study area. The other five soil heavy metals (Pb, Cu, Cd, Cr, and As) were in safe levels of soil environmental quality assessment classification standard. The average weighted pollutant index was 0.54, indicating good soil environmental quality. Most parts of the region was in safe level. An exponential smoothing forecasting model was established based on the time series characteristics of the heavy metal pollution in the agricultural producing area to predict heavy metal pollution. The analytical results showed that the prediction accuracy of the models was approximately 90%, indicating that the exponential smoothing model can fit the variation of heavy metal pollution.