Collaborative Research: Unraveling Orographic Precipitation Patterns by Combined Hydrologic and Atmospheric Analysis

我们将以加利福尼亚的南内华达山脉为重点，使用分布的雪和流量测量（可追溯到20世纪20年代）来评估每个盆地每年和每次风暴时间尺度上的降水量。在贝叶斯框架中，通过一系列模型结构的迭代水文模拟，我们将确定哪种降水量和梯度（以及不确定性）最适合现有的流量和积雪测量。这种方法将为没有高海拔雨量计的时间和地点提供基于地面的高海拔降水估计，并使我们能够评估长期变化。然后，我们将使用这些新的基于水文的降水数据集作为大气模式性能的基准，并将这些数据与使用不同边界条件和微物理方案生成的14年区域大气模式（WRF）模拟集合进行比较。这将使我们能够评估地形降水年模式不确定性的来源。准确的降水输入，无论是总量还是随海拔的变化，都是模拟山区河流的关键。然而，总的来说，就降水测量而言，这些地区的采样严重不足，而那些确实存在的测量仪是出了名的不可靠。大气模式可以预测降水率和分布。然而，这些大气模型的发展和改进一直受到缺乏直接降水测量的阻碍，这首先给水文学家带来了问题。我们从这些高海拔地区获得的最可靠的测量结果（分布的河流流量和雪水当量）是估计降水历史时空分布的有用线索。我们的研究方法整合了气象学和水文学，利用水文学作为更好地理解气象学的工具，并提供了一个长期的基准，可用于改进我们的预测和推进这两个领域的科学。这项工作将提高对单个风暴和这些风暴如何在整个季节累积的山区降水的理解。这些知识对于预测短期洪水、季节性水资源和长期气候敏感性至关重要。这些努力将集中在加州内华达山脉南部，那里的短期和长期预测历来比其他地区的准确性低，但这些预测对价值超过300亿美元的圣华金河谷农业至关重要。研究结果也将有助于大气模式的发展，使这些模式可以更成功地用于预测全球山区降水。

Focusing on the Southern Sierra Nevada, California, we will use distributed snow and streamflow measurements (which date back to the 1920s) to assess what precipitation must have been for each basin on annual and storm-by-storm time-scales. Using iterative hydrologic simulations across a range of model structures in a Bayesian framework, we will determine which precipitation amounts and gradients (and with what uncertainty) best fit available measurements of streamflow and snow accumulation. This methodology will provide a ground-based estimate of high-elevation precipitation for times and locations where no high-altitude rain gauges are available and will allow us to assess long-term change. We will then use these new hydrologic-based precipitation datasets as a benchmark for atmospheric model performance, and will compare these with an ensemble of 14-year regional atmospheric model (WRF) simulations generated using different boundary conditions and microphysics schemes. This will allow us to assess sources of model uncertainty in annual patterns of orographic precipitation. Accurate precipitation inputs, both total amounts and changes with elevation, are critical to model streamflow in mountainous regions. However, in general these regions are grossly under-sampled in terms of precipitation measurements, and those gauges that do exist are notoriously unreliable. Atmospheric models can predict precipitation rates and distributions. However, development and improvement of these atmospheric models has been hindered by the lack of direct precipitation measurements that caused hydrologists problems in the first place.Our most reliable measurements from these high-altitude areas (distributed streamflow and snow water equivalent) are useful clues for estimating historic spatial and temporal distributions of precipitation. Our research approach integrates both meteorology and hydrology, using hydrology as a tool to better understand meteorology, and providing a long-term benchmark that can be used to improve our forecasts and advance science in both fields. This work will improve understanding of mountain precipitation for both individual storms and how those storms aggregate over an entire season. This knowledge is crucial for forecasting short-term floods, seasonal water resources, and long-term climate sensitivity. Efforts will focus on the Southern Sierra Nevada, California, where short-term and long-term forecasts have historically had lower accuracy than other regions, but where those same forecasts are critical for the San Joaquin Valley agriculture industry, valued at over $30 billion. The research results will also contribute to atmospheric model development, such that these models can be used more successfully to predict mountain precipitation worldwide.