Abstract We use the Budyko framework to examine streamflow changes due to climate variation and human activities in the Mauvais Coulee Basin, a 1002 km2 sub-basin that contributes streamflow to Devils Lake, a terminal lake that has experienced catastrophic flooding over the past three decades. The basin lies within the northern glaciated prairies, contains innumerable wetland complexes that follow fill-spill hydrological principles, has undergone extensive land use change to support dryland farming, and has experienced a drought to deluge hydroclimatic transition over the last century. We implement the Budyko framework to partition the runoff increase to the lake between its climatic or human components. Initial results attribute 45% of the streamflow increase to climate factors, and 55% to human activities. Implementation of the Budyko framework to examine streamflow partitioning, however, rests upon a series of assumptions that are usually not directly verified. We then examine each Budyko framework assumption to determine its suitability for our study site and study period. None of the required assumptions are fully satisfied. Our study site and study period encompass two distinct hydroclimatic modes, contain no clear single inflection point separating baseline and altered periods, experience root zone soil moisture storage changes during the study period, and exhibit a non-stationary precipitation-streamflow relationship during the transitional period between the two hydroclimatic modes. Streamflow attribution produced by the standard application of the Budyko framework is not applicable to our study site or study period. We recommend that local-scale investigations of streamflow partitioning based upon the Budyko framework be more intentional in verifying model assumptions as a standard practice in implementing the method.

中文翻译：

---

在干旱到洪水条件下的北部冰川流域中使用 Budyko 框架划分水流

摘要 我们使用 Budyko 框架来研究 Mauvais Coulee 盆地由于气候变化和人类活动而导致的水流变化，这是一个 1002 平方公里的子流域，为魔鬼湖提供水流，魔鬼湖是一个在过去三十年中经历了灾难性洪水的终端湖。该盆地位于北部冰川草原内，包含无数遵循溢流水文原理的湿地复合体，经历了广泛的土地利用变化以支持旱地农业，并在上个世纪经历了干旱导致水文气候转变。我们实施了 Budyko 框架来划分湖泊的气候或人类成分之间的径流增加。初步结果将 45% 的流量增加归因于气候因素，55% 归因于人类活动。但是，用于检查流流分区的 Budyko 框架的实现依赖于通常未直接验证的一系列假设。然后，我们检查每个 Budyko 框架假设，以确定其对我们的研究地点和研究时期的适用性。没有一个要求的假设完全得到满足。我们的研究地点和研究期包含两种不同的水文气候模式，不包含区分基线和改变期的明确单一拐点，在研究期间经历根区土壤水分储存变化，并在过渡期表现出非平稳的降水-流量关系在两种水文气候模式之间。Budyko 框架的标准应用产生的流流归属不适用于我们的研究地点或研究时期。