Water management agencies seek the next generation of modeling tools for planning and operating river basins. Previous site‐specific models such as U.S. Bureau of Reclamation's (USBR) Colorado River Simulation System and Tennessee Valley Authority's (TVA) Daily Scheduling Model have become obsolete; however, new models are difficult and expensive to develop and maintain. Previous generalized river basin modeling tools are limited in their ability to represent diverse physical system and operating policy details for a wide range of applications. RiverWare(tm), a new generalized river basin modeling tool, provides a construction kit for developing and running detailed, site‐specific models without the need to develop or maintain the supporting software within the water management agency. It includes an extensible library of modeling algorithms, several solvers, and a rich “language” for the expression of operating policy. Its point‐and‐click graphical interface facilitates model construction and execution, and communication of policies, assumptions and results to others. Applications developed and used by the TVA and the USBR demonstrate that a wide range of operational and planning problems on widely varying basins can be solved using this tool.

[1] Stochastic disaggregation models are used to simulate streamflows at multiple sites preserving their temporal and spatial dependencies. Traditional approaches to this problem involve transforming the streamflow data of each month and at every location to a Gaussian structure and subsequently fitting a linear model in the transformed space. The simulations are then back transformed to the original space. The main drawbacks of this approach are (1) transforming marginals to Gaussian need not lead to the correct multivariate distribution particularly if the dependence across variables is nonlinear, and (2) the number of parameters to be estimated for a traditional disaggregation model grows rapidly with an increase in space or time components. We present a K-nearest-neighbor approach to resample monthly flows conditioned on an annual value in a temporal disaggregation or multiple upstream locations conditioned on a downstream location for a spatial disaggregation. The method is parsimonious, as the only parameter to estimate is K (the number of nearest neighbors to be used in resampling). Simulating space-time flow scenarios conditioned upon large-scale climate information (e.g., El Niño-Southern Oscillation, etc.) can be easily achieved. We demonstrate the utility of this methodology by applying it for space-time disaggregation of streamflows in the Upper Colorado River basin. The method appropriately captures the distributional and spatial dependency properties at all the locations.

This paper presents a lag-1 modified K-nearest neighbor ͑K-NN͒ approach for stochastic streamflow simulation. The simulation at any time t given the value at the time t − 1 involves two steps: ͑1͒ obtaining the conditional mean from a local polynomial fitted to the historical values of time t and t − 1, and ͑2͒ then resampling ͑i.e., bootstrapping͒ a residual at one of the historical observations and adding it to the conditional mean. The residuals are resampled using a probability metric that gives more weight to the nearest neighbor and less to the farthest. The "residual resampling" step is the modification to the traditional K-NN time-series bootstrap approach, which enables the generation of values not seen in the historical record. This model is applied to monthly streamflow at the Lees Ferry stream gauge on the Colorado River and is compared to both a parametric periodic autoregressive and a nonparametric index sequential method for streamflow generation, each widely used in practice. The modified K-NN approach is found to exhibit better performance in terms of capturing the features present in the data.

[1] As multicentury records of natural hydrologic variability, tree ring reconstructions of streamflow have proven valuable in water resources planning and management. All previous reconstructions have used parametric methods, most often regression, to develop a model relating a set of tree ring data to a target hydrology. In this paper, we present the first development and application of a K nearest neighbor (KNN) nonparametric method to reconstruct naturalized annual streamflow ensembles from tree ring chronology data in the Upper Colorado River Basin region. The method is developed using tree ring chronologies from the period 1400-2005 and naturalized streamflow from the period 1906-2005 at the important Lees Ferry, Arizona, gauge on the Colorado River to develop annual streamflow ensembles for this gauge for the 1400-1905 period. The proposed KNN algorithm was developed and tested using cross validation for the overlap period, i.e., the contemporary observed period for which both the tree ring and streamflow data are available . The cross-validated streamflow reconstructions for the selected contemporary period compare very well with the observed flows and also with published parametric streamflow reconstructions for this gauge. The proposed nonparametric method provides an ensemble of streamflows for each year in the paleohydrologic reconstruction period and, consequently, a more realistic asymmetric confidence interval than one obtained through most parametric approaches. Also, the K nearest neighbors are obtained only from the tree ring chronology data, and thus, the method can be used to reconstruct structured and even nonnumerical data for use in water resources modeling.