Impact of Forest Service Activities on Stream Flow

Home Browse Search

TroendlelNankervis/Porth Page27 5122/2003 Appendix A The Evolution of Hydrologic Thinking and the Models That Resulted In the late 1970's, the WRENSS hydrologic model (Troendle and Leaf, 1980) was developed as part of a lazger nomographic procedure to assess Water Resources Evaluation of Non-Point Sources in Silviculture (EPA 1980). The WRENSS hydrologic model (Troendle and Leaf 1980) was revised to reflect current hydrologic thinking and used in this effort to simulate both the historical changes in stream flow from the Forest Service Lands in the North Platte River and to simulate the impact of recent (1997 to 2017) management activities on that stream flow. The latest revision is also somewhat different than the version of the model used in the earlier North Platte analysis (Troendle and Nankervis 2000). T'he current WRENSS hydrologic model is a stand-alone version, for the Rocky Mountain Region (Hydrologic Regian 4, Troendle and Leaf 1980) that has been modified from earlier application by Troendle and Nankervis (2000) primarily in the manner in which precipitation inputs aze handled. The Evapotranspiration (ET) Modifier Coefficients that aze used to adjust evapotranspiration following forest density reduction ha.ve also been changed but this has a lesser impact on response. The most significant alteration in the current version, in terms of total stream flow simulated, results from a modification in how precipitation is managed. In previous versions of the WRENSS hydrologic model, the estimates of winter and spring precipitation input to the model were adjusted upward following timber harvest, to reflect the difference in accumulation observed to occur as stand density is reduced. Originally (Troendle and Leaf 1980), the increase in accumulation was thought to be the result of a redistribution process resulting in differentia.l deposition between forest and open. Since then, the conclusion has been dra.wn that the increase in snow pack deposition represents a reduction in what would have been an interception (vaporization) loss. In either case, the estimate of precipitation input to the model was increased to reflect the differential in accumulation in the harvested azea. Depending on whether or not the estimate of precipitation input to the model was an estimate of the gross (above canopy) or net (below canopy) precipitation input to the site, consistently increasing the estimate of precipitation to reflect the differences in accumulation observed under the harvested condition could result in over estimating net precipitation entering the site. In the version of the WRENSS hydrologic model used in this ana.lysis, it is assumed that the estimate of precipitation input to the model represents the gross precipitation entering the top of the canopy and therefore subject to canopy interception and subsequent evaporative losses. The WRENSS Hydrologic model has been modified such that, depending on stand density, the precipitation estimate input to the model is adjusted downward to account for net interception losses that may occur from tha.t portion intercepted in the canopy during the winter and spring. The interception functions in the current revision of the WItENSS hydrologic model assume gross precipitation to be synonymous with the amount of precipitation deposited in a forest opening (see Sturges 1984, Schmidt and Troendle 1992) with lesser amounts deposited under stands of increasing density. The interception functions (Figure 1) are the reciprocal of the interception functions used in the previous version of WRENS5 hydrologic model (see Nankervis and Troendle 2000, Figures 13 - 15). Continuity of the current version of WRENSS with the earlier version, with respect to the relative differences in snow pack accumulation between forest and open as a result of a net reduction in interception loss, is maintained but the resulting estimate of net precipitation to the site is reduced by the amount of winter/spring