Challenges in Identifying Hydrometeorologically Homogenous Regions for use in Hydrologic Modeling with Uncertainty, Leveraging Forty Years of the Analysis of Record for Calibration (AORC) Dataset

Stochastic Storm Transposition (SST) is a methodology used to move observed precipitation associated with a storm event from its original location to multiple, randomly selected alternate locations within a climatologically homogeneous region for the purpose of probabilistic hydrologic modeling. SST can be a particularly useful tool in regions where precipitation frequency estimates are unavailable due to a lack of existing or up-to-date frequency data. The SST method also provides a valuable alternative to traditional storm estimation methods, which often utilize synthetic, elliptical spatial patterns and uniform or idealized temporal distributions, by instead moving observed storms with both realistic temporal and spatial rainfall patterns.

The challenges for developing datasets for use with SST have been eased due to advancements in computational power combined with several decades of remotely sensed quantitative precipitation estimates, like the Analysis of Record For Calibration (AORC) dataset, which allow for automated storm identification and cataloging. A fundamental requirement for the SST process is the delineation of a transposition region, defined as a climatologically homogenous region in which any observed storm contained within could plausibly be moved to any other location in that region. Developing transposition regions remains a subjective process in which meteorological knowledge is combined with gridded climatological variables to inform delineation. The most common climatological variable for informing decisions is mean annual (seasonal) precipitation, however, mean and maximum dewpoints, prevailing wind conditions, distance to major moisture sources, and historical tropical cyclone tracks can also provide valuable information; combination of these variables is often used depending on the specific geography of an area of interest.

The process of region development is further complicated when considering large watersheds that are very unlikely to be climatology homogenous themselves. For example, mountainous basins in the western United States can have nearly 10,000 foot differences in elevation from the highest to lowest points, and drastically different precipitation patterns based on elevation and aspect. Complexity continues to compound when considering basins where the majority of annual precipitation is frozen and acts to build snowpack, rather than runoff.

Increasing objectivity and reducing uncertainty are obvious goals of transposition region development. Looking towards the future, statistical testing of domains via ratio distributions and storm typing may provide that. Storm typing can further refine identified storms within a transposition region by classifying them as synoptic-scale tropical or extra-tropical cyclonic events, mesoscale complexes, or localized convection, and ensuring that all identified storms are realistic for transposition.

This presentation will discuss the process of developing climatologically homogeneous storm transposition regions for watersheds across the Western United States, the complications and limitations that come with each transposition region due to spatially varied topography and hydrology, validation and verification efforts to objectively assess storm transposition region suitability, the subsequent hydrologic modeling done for a watershed, and continued efforts to improve the SST process.