This report presents the results of a three-step evaluation of the stream-gaging program in Wisconsin. First, data uses and funding sources were identified for the 89 continuous-record gaging stations operated during the 1984 water year. Next, alternative methods of streamflow estimation were examined for three stations. A flow-routing model was used for two of the stations and a statistical model was used for the third. The modeled discharges did not compare well enough with observed discharges to warrant elimination of any of the stations. Finally, an optimization model was used to assess the cost effectiveness of the stream-gaging process.
The annual budget, in 1984 dollars, for operating the 89 continuous-record gaging stations and 65 additional stations is $557,3000. Based on a Kalman-filter analysis, the theoretical average standard error of instantaneous discharge associated with the current practice of visiting the stations is 13.8%. This overall level of accuracy could be maintained with a budget of $518,6000 if stream-gaging activities were redistributed in an optimal fashion among the gaging stations. For the current budget, the theoretical average standard error would be reduced to 10.1% if the network is operated in an optimal fashion. Furthermore, the average standard error would be reduced to 7.3% if all missing record is eliminated and the network is operated optimally.
A minimum budget of $510,000 is required to operate the program; a budget less than this does not permit proper service and maintenance of the gaging stations. At this minimum budget, the theoretical average standard error of instantaneous discharge is 14.4%. The maximum budget analyzed was $650,000 and resulted in an average standard of error of instantaneous discharge of 7.2%.
