Optimal Field Management for Regional Water Quality Planning

Babak Negahban, C.B. Moss, J.W. Jones, J. Zhang, W.D. Boggess, K.L. Campbell

ABSTRACT:
The process of regional planning can be thought of as an attempt to achieve certain regional goals by implementing local modifications in activities. The development of computer software and hardware has enabled the efficient organization of an extensive amount of local (small scale) information which can be combined into a regional (large scale) plan. There is an exceedingly large number of decision choices to be made at the local level because of the number of local fields (land units), the spatial variability in fields, and the combinations of land use and management practices that are possible. Optimization is a tool which can sift through the numerous combinations of local choices to pick those which, when combined, will produce an optimum plan which best meets regional goals within the constraints imposed on combinations of activities.

The Lake Okeechobee Agriculture Decision Support System (LOADSS) is a GIS-based system which allows planners to assign any of over 100 management practices on any of over 8000 fields in the Lake Okeechobee watershed in South Florida with the goal of controlling excessive nutrient runoff into the lake. Since the number different combinations of management practices on fields are exceedingly large (> 8000100), an optimization module was added to LOADSS to automatically assign practices to fields in order to best meet stated regional goals. Planners can use a mouse and menu-driven user interface to develop optimization formulations incorporating any of 40 environmental, economic, import or export attributes available in the LOADSS database.

In order to test the effectiveness of using optimization algorithms in LOADSS, a pilot basin was selected for optimal regional planning. Eight optimization formulations were selected to be tested on the pilot basin of which six achieved optimal results. Of the six optimal solutions that were obtained for the pilot basin, all but one were able to improve on the LOADSS base plan for the basin in both economic and environmental terms.

LOADSS requires the ARC/INFO 6.1 (GIS) and GAMS (optimization) software and executes on SPARC architecture workstations. 

Babak Negahban
Senior GIS Analyst
Breedlove, Dennis & Associates, Inc.
P.O. Bos 720037
Orlando, Florida 32872-0037