Empirical models of annual post-fire erosion on mulched and unmulched hillslopes
Empirical models of annual post-fire erosion on mulched and unmulched hillslopes
复制标题
DOI：
10.1016/j.catena.2017.12.029
复制
发表时间：
2018-04
期刊：
影响因子：
6.2
通讯作者：
Sarah Schmeer;S. Kampf;L. Macdonald;Joshua Hewitt;C. Wilson
中科院分区：
文献类型：
--
作者：
Sarah Schmeer;S. Kampf;L. Macdonald;Joshua Hewitt;C. Wilson
侵蚀是野火后土地管理的主要问题之一。本研究研究了对 2012 年科罗拉多州北部 High Park 火灾后山坡规模侵蚀的控制，开发了用于预测火灾后沉积物产量的简单经验模型，并评估了附近几场火灾的模型性能。从 2013 年到 2015 年，我们收集了 29 个收敛山坡的地表覆盖、降雨量、地形和产沙量测量数据；其中八个山坡施用了不同数量的覆盖物以减少侵蚀。根据这些数据，我们研究了年沉积物产量与三类预测变量（地面覆盖、降水和地形）之间的相关性。裸土百分比是对沉积物产量的最大控制因素，其次是降雨量变量。产沙量普遍随流路长度的增加而降低，但相关性较弱。每个经验模型都用三个变量预测沉积物产量：裸土百分比、一个降水变量和一个地形变量。使用不同的降水量和地形变量组合（Nash-Sutcliffe 系数 0.70-0.84），模型对于 High Park Fire 具有相似的精度。将年沉积物产量预测为裸土百分比、6 月至 10 月降水量和最大水流路径长度函数的经验模型在应用于同一地区的其他火灾时表现各异，对个别火灾的预测范围从差到好，所有火灾综合的纳什-萨克利夫系数为 0.26-0.32。这些测试显示了将经验模型应用于研究区域火灾的一些希望，但需要进一步的模型测试来确定可以应用该模型的条件范围。
Erosion is one of the primary land management concerns following wildfire. This study examines controls on post-fire hillslope-scale erosion for the 2012 High Park Fire in northern Colorado, develops simple empirical models for predicting post-fire sediment yields, and evaluates model performance on several nearby fires. From 2013 to 2015 we collected ground cover, rainfall, topographic, and sediment yield measurements from 29 convergent hillslopes; eight of these hillslopes had varying amounts of mulch applied to reduce erosion. From these data we examined correlations between annual sediment yield and three categories of predictor variables (ground cover, precipitation, and topography). Percent bare soil was the single largest control on sediment yield, followed by rainfall variables. Sediment yield generally decreased with flow path length, but the correlation was weak. The empirical models each predicted sediment yield with three variables: percent bare soil, one precipitation variable, and one topographic variable. The models had similar accuracy for the High Park Fire using varying combinations of precipitation and topographic variables (Nash-Sutcliffe coefficients 0.70–0.84). An empirical model predicting annual sediment yields as a function of percent bare soil, June–October precipitation, and the maximum flow path length had variable performance when applied to other fires in the same region, with predictions ranging from poor to good for individual fires and Nash-Sutcliffe coefficients of 0.26–0.32 for all fires combined. These tests show some promise for applying the empirical model to fires in the study region, but further model testing is needed to determine the range of conditions under which the model can be applied.