Impact of Forest Service Activities on Stream Flow

Home Browse Search

Troendle/Nankervis/Por:h Page 18 5/22/2003 minimal impact on stream flow and the; increases in stream flow resulting from those activities are not offsetting the natural decline. A.s a point of reference, we used the 1992 Environmental Protection Agency (EPA) National Land Classification (NLC) data to determine the extent of forested area in the North Platte basin in Wyoming and Colorado. By intersecting the North Platte boundary on the I?PA NLC layer we determined that approximately 1.64 million acres of forestland occur in the Wyoming and Colorado portion of the watershed. Approximately 62 percent of that forestlancl is administereci by the National Forest Systems. Development of a Reference Sysitem to Docurment Change The second major task in this effort irivolved identifying a series of stream flow, snow pack accumulation, and precipitation monitoring sites that would be useful as reference pairs for the future documentation of' measurable lorig-term changes in stream flow. The changes in flow of concern are those that might occur as a r-esult of NFS Forest Management activities, or other land and water use changes, the initial intent was to select 6 or so long-term stream gauges that could be used to index water yield and then identify long-term snow course or precipitation records that could be used, in a double-mass analysis to document future change. Critical to this procedure is the ability to identi.fy long-term monitoring sites with stable records suitable for such an analysis. The process, detailed 'ui Appendix C is summazized as follows. The analysis was sequential in na.ture. F'irst we needed to identify a series of stream gauges that could be used to docw?nent flow changes. Since it was a flow change that we wished to document, we then needed to identify stable climatic data that could be used in the double-mass comparison. It was assuined that the cliimatic data would not be subject to change and therefore changes in the double-mass relationships could be inferred to be the result of flow changes. This assumption must be evabua.ted as part of the analysis. Stream Gauge Selection First, we identified the suite of strearn gauges available for analysis, and critical selection criteria, in addition to staLbility, were that the gauge had to have a minimum of 25 years of record prior to 1997 and the gautge must currenl:ly be in operation. Stream gauge data was obtained from the USGS website: httD://waterdata.usjzs;gov/nwis. We found that only 194 of the 1152 stream gauges in Colorado and 96 of the 656 stream gauges in Wyoming had the combination of 25 years of record and were active through 1997. Of those, only 27 stream gauges in Colorado and 20 in Wyoming were located on either the North or South Platte River. Daily stream flow data for the entire period of record was downloaded for each of the 47 sites. Since all of the stream gauges aze not maintaine:d yeaz azound and because observed changes in flow due to vegetation changes have been obset•ved to occur diuing the May to September period, only seasonal flows were utilized. Seasona.l ilow is that measured during the time period April 15 to September 30 each year. Seasonal records for each ,gauging site were assessed for missing data, prior to continuing the analysis. ]:f one or two days of record were missing near the beginning or the end of the seasonal record the value was estimated as the average of the discharge on the preceding and next day, assuming the two values were similaz and available to calculate a mean flow. If more days were missing or data was missing during the high flow period, the data for tha.t year was deleted. Of the 3086' years of recori3 available for the 47 strea.m gauges selected for further