Integrated spatial decision support system for precision agriculture.

摘要

Excessive application of plant nutrients and pesticides on agricultural land has resulted in both environmental degradation and economic loss to the farming community. Agricultural non-point source pollution was cited as the primary source of the water quality problems in many areas of the United States. Environmental concerns resulting from agricultural non-point source pollution has placed demands on farmers and ranchers to implement the best management practices (BMPs) to reduce the delivery of pollutants to streams and aquifers. Precision agriculture, a relatively recent crop production and agricultural management strategy holds great promise to minimize environmental pollution while to maximize economic productivity and profitability. It has benefited from rapidly evolving geospatial information technologies, such as global positing systems (GPS), geographic information systems (GIS), remote sensing (RS), and electronic sensors and “intelligent” controllers. However, the complexity of making routine, coherent, and cost-effective farm management decisions presents a formidable challenge to farmers. What is lacking in precision agriculture is an analytical tool that integrates these component technologies with biophysical and economic models for tactical, strategic, and policy-level decision make. In this dissertation, a decision support system called IDSSPA is developed to include modules for evaluating crop yield and chemical losses in response to site-specific management of agricultural inputs. Using this system, not only can users store, visualize, manipulate, and analyze spatial/non-spatial field experiment data, but they also can do various simulations through the easy-operated biophysical models, which take field spatial variability into account. In the system, the functionalities of the traditional models and analysis methods have been enhanced by coupling them with each other and with ArcView GIS. Uniquely designed GIS-based interfaces enable the lumped biophysical models to incorporate and represent field spatial variability. Statistical and data mining tools are also included in the system to improve analysis of field measured data and to further enhance interpretation of model simulation results. Other components incorporated into the system are as follows: The CERES-Maize plant growth model seamlessly integrated with RZWQM to provide an alternative phonologically based model for predicting growth and yield of maize (corn), and several tools for evaluating economic and ecologic risks of precision agriculture implementation. The application examples indicated that IDSSPA is a useful research and decision make tool for precision agriculture at field and watershed scales.