• Advances in seasonal forecasting for water management in Arizona: a case study of the 1997-98 El Niño

      Pagano, Thomas; Hartmann, Holly; Sorooshian, Soroosh; Bales, Roger; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1999-11)
      This 1997-98 El Niño provided a unique opportunity for climate information and forecasts to be utilized by water management agencies in the Southwestern U.S. While Arizona has experienced high streamflow associated with previous El Niño events, never before had an event of such magnitude been predicted with advance warning of several months. Likewise, the availability of information, including Internet sources and widespread media coverage, was higher than ever before. Insights about use of this information in operational water management decision processes are developed through a series of semi -structured in -depth interviews with key personnel from a broad array of agencies responsible for emergency management and water supply, with jurisdictions ranging from urban to rural and local to regional. The interviews investigate where information was acquired, how it was interpreted and how it was incorporated into specific decisions and actions. The interviews also investigate agency satisfaction with the products available to them, their operational decisions, and intentions to utilize forecast products in the future. Study findings lead to recommendations about how to more effectively provide intended users of forecasts with information required to enact mitigation measures and utilize opportunities that some climatic events present. The material presented in this report is primarily based on the Masters Thesis of Thomas Pagano.

      Hendrickson, Jene Diane,1960-; Sorooshian, Soroosh; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1987-05)
      In the past, derivative-based optimization algorithms have not frequently been used to calibrate conceptual rainfall -riff (CRR) models, partially due to difficulties associated with obtaining the required derivatives. This research applies a recently- developed technique of analytically computing derivatives of a CRR model to a complex, widely -used CRR model. The resulting least squares response surface was found to contain numerous discontinuities in the surface and derivatives. However, the surface and its derivatives were found to be everywhere finite, permitting the use of derivative -based optimization algorithms. Finite difference numeric derivatives were computed and found to be virtually identical to analytic derivatives. A comparison was made between gradient (Newton- Raphsoz) and direct (pattern search) optimization algorithms. The pattern search algorithm was found to be more robust. The lower robustness of the Newton-Raphsoi algorithm was thought to be due to discontinuities and a rough texture of the response surface.

      Duan, Qingyun; Sorooshian, Soroosh; Ibbitt, Richard P.; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1987-08)
      A new Maximum Likelihood Criterion (MLE) suitable for data which are recorded at unequal time intervals and contain auto-correlated errors is developed. Validation of the new MLE criterion has been carried out both on a simple two - parameter reservoir model using synthetical data and on a more complicated hillslope model using real data from the Pukeiti Catchment in New Zealand. Comparison between the new MLE criterion and the Simple Least Squares (SLS) criterion reveals the superiority of the former over the latter. Comparison made between the new MLE and the MLE for auto-correlated case proposed by Sorooshian in 1978 has shown that both criteria would yield results with no practical difference if equal time interval data were used. However, the new MLE can work on variable time interval data which provide more information than equal time interval data, and therefore produces better visual results in hydrologic simulations.

      Winchell, Michael; Gupta, Hoshin Vijai; Sorooshian, Soroosh; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1997)
      Runoff generation has been shown to be very sensitive to precipitation input. With the use of precipitation estimates from weather radar, errors are introduced from both the transformation from reflectivity to precipitation rate and the spatial and temporal aggregation of the radar product. Currently, a significant degree of uncertainty exists in the accuracy of radar-based precipitation estimates. When uncalibrated or poorly calibrated radar products were used as input to a rainfall-runoff model, the resulting predicted runoff varied severely from the runoff generated using well-calibrated radar products. Another source of uncertainty, errors in the precipitation system structure due to aggregation in time and space, has also been shown to affect runoff generation. This study focuses on separating the primary runoff- generating mechanisms -- infiltration excess and saturation excess -- to assess their responses to variable precipitation inputs individually. For the case of saturation excess runoff, there was minimal sensitivity due to temporal aggregation of the precipitation; however, there was considerable sensitivity to spatial aggregation. For the case of infiltration excess runoff, temporal and spatial aggregation of the precipitation significantly reduced the amount of runoff produced. The magnitudes of these runoff reductions varied between storms and showed a high degree of dependence on storm characteristics, particularly the maximum precipitation intensity.

      Humes, Karen Sue; Sorooshian, Soroosh; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1993)
      The overall topic of the research described in this dissertation was the partitioning of available energy at the Earth's surface into sensible and latent heat flux, with an emphasis on the development of techniques which utilize remotely sensed data. One of the major objectives was to investigate the modification of existing techniques, developed over agricultural surfaces, to "natural" ecosystems (i.e., non -agricultural vegetation types with variable and incomplete canopy cover). Ground -based measurements of surface fluxes, vegetation cover, and surface and root -zone soil moisture from the First ISLSCP (International Land Surface Climatology Program) Field Experiment (FIFE) were used to examine the factors controlling the partitioning of energy at ground stations with contrasting surface characteristics. Utilizing helicopter -based and satellite -based data acquired directly over ground -based flux stations at the FINE experimental area, relatively simple algorithms were developed for estimating the soil heat flux and sensible heat flux from remotely sensed data. The root mean square error (RMSE) between the sensible heat flux computed with the remotely sensed data and the sensible heat flux measured at the ground stations was 33 Wm 2. These algorithms were then applied on a pixel -by -pixel basis to data from a Landsat -TM (Thematic Mapper) scene acquired over the FIFE site on August 15, 1987 to produce spatially distributed surface energy- balance components for the FIFE site. A methodology for quantifying the effect of spatial scaling on parameters derived from remotely sensed data was presented. As an example of the utility of this approach, NDVI values for the 1,IFE experimental area were computed with input data of variable spatial resolution. The differences in the values of NDVI computed at different spatial resolutions were accurately predicted by an equation which quantified those differences in terms of variability in input observations.
    • Evaluation of national weather service ensemble streamflow predictions (ESP) for the Colorado river basin

      Franz, Kristie J.; Hartmann, Holly C.; Sorooshian, Soroosh; Bales, Roger; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 2003-01)
      The National Weather Service (NWS) developed the Ensemble Streamflow Prediction (ESP) system to generate probabilistic water supply forecasts that consist of an ensemble of streamflow traces (scenarios) conditioned on the initial states of the basin. Uncertainty information available from ESP may be more useful to risk-based decision makers than what is currently provided by linear regression forecasts. Because the use of ESP has been limited to date, there are few operational forecasts available for verification. Therefore, it was necessary to generate simulated operational ESP forecasts to test the forecasting procedure's potential to enhance current forecasting techniques. Simulated historical forecasts were generated for 14 forecast points in the Colorado River basin. The median and best forecast traces were analyzed as representations of ESP deterministic forecasts. General scalar statistics were used to evaluate these traces. The probability information contained in the entire ensemble was analyzed using probabilistic and conditional verification methods. It was found that the information contained in the median trace is limited and that choosing one trace is not the optimal use of ESP forecast information. ESP provides a probabilistic forecast that performs better than a probabilistic forecast based on climatology. In addition, ESP can provide accurate information about the magnitude of future streamflow discharge even at lead times of up to seven months. With shorter lead times (2-3 months), the forecasts become more informative and accurate.
    • Improving the Reliability of Compartmental Models: Case of Conceptual Hydrologic Rainfall-Runoff Models

      Sorooshian, Soroosh; Gupta, Vijai Kumar; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1986-08)
    • Investigation of the national weather service soil moisture accounting models for flood prediction in the northeast floods of january 1996

      Hogue, Terri S.; Sorooshian, Soroosh; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1999-10)
      Extensive flooding occurred throughout the northeastern United States during January of 1996. The flood event cost the lives of 33 people and over a billion dollars in flood damage. Following the `Blizzard of `96 ", a warm front moved into the Mid-Atlantic region bringing extensive rainfall and causing significant melting and flooding to occur. Flood forecasting is a vital part of the National Weather Service (NWS) hydrologic responsibilities. Currently, the NWS River Forecast Centers use either the Antecedent Precipitation Index (API) or the Sacramento Soil -Moisture Accounting Model (SAC-SMA). This study evaluates the API and SAC -SMA models for their effectiveness in flood forecasting during this rain -on -snow event. The SAC -SMA, in conjunction with the SNOW-17 model, is calibrated for five basins in the Mid -Atlantic region using the Shuffled Complex Evolution (SCE-UA) automatic algorithm developed at the University of Arizona. Nash-Sutcliffe forecasting efficiencies (Ef) for the calibration period range from 0.79 to 0.87, with verification values from 0.42 to 0.95. Flood simulations were performed on the five basins using the API and calibrated SAC-SMA model. The SAC-SMA model does a better job of estimating observed flood discharge on three of the five study basins, while two of the basins experience flood simulation problems with both models. Study results indicate the SAC-SMA has the potential for better flood forecasting during complex rain-on-snow events such as during the January 1996 floods in the Northeast.
    • A multi-step automatic calibration scheme (MACS) for river forecasting models utilizing the national weather service river forecast system (NWSRFS)

      Sorooshian, Soroosh; Gupta, Hoshin; Hogue, Terri S.; Holz, Andrea; Braatz, Dean; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1999-10)
      Traditional model calibration by National Weather Service (NWS) River Forecast Center (RFC) hydrologists involves a laborious and time -consuming manual estimation of numerous parameters. The National Weather Service River Forecasting System (NWSRFS), a software system used by the RFCs for hydrologic forecasting, includes an automatic optimization program (OPT3) to aid in model calibration. The OPT3 program is not used operationally by the majority of RFC hydrologists who perform calibration studies. Lack of success with the traditional single - step, single-criterion automatic calibration approach has left hydrologists more comfortable employing a manual step-by-step process to estimate parameters. This study develops a Multistep Automatic Calibration Scheme (MACS), utilizing OPT3, for the river forecasting models used by the RFCs: the Sacramento Soil Moisture Accounting (SAC-SMA). and SNOW-17 models. Sixteen parameters are calibrated in three steps, replicating the progression of manual calibration steps used by NWS hydrologists. MACS is developed by minimizing different objective functions for different parameters in a step -wise manner. Model runs are compared using the MACS optimized parameters and the manually estimated parameters for six basins in the North Central River Forecast Center (NCRFC) forecast area. Results demonstrate that the parameters obtained via the MACS procedure generally yield better model performance than those obtained by manual calibration. The MACS methodology is a time-saving approach that can provide prompt model forecasts for NWS watersheds.
    • A multiobjective global optimization algorithm with application to calibration of hydrologic models

      Yapo, Patrice O.; Gupta, Hoshin Vijai; Sorooshian, Soroosh; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1997-02)
      This report presents a new multiple objective optimization algorithm that is capable of solving for the entire Pareto set in one single optimization run. The multi-objective complex evolution (MOCOM-UA) procedure is based on the following three concepts: (1) population, (2) rank-based selection, and (3) competitive evolution. In the MOCOM-UA algorithm, a population of candidate solutions is evolved in the feasible space to search for the Pareto set. Ranking of the population is accomplished through Pareto ranking, where all points are successively placed on different Pareto fronts. Competitive evolution consists of selecting subsets of points (including all worst points in the population) based on their ranks and moving the worst points toward the Pareto set using the newly developed multi-objective simplex (MOSIM) procedure. Test analysis on the MOCOM-UA algorithm is accomplished on mathematical problems of increasing complexity and based on a bi-criterion measure of performance. The two performance criteria are: (1) efficiency, as measured by the ability of the algorithm to converge quickly, and (2) effectiveness, as measured by the ability of the algorithm to locate the Pareto set. Comparison of the MOCOM-UA algorithm against three multi-objective genetic algorithms (MOGAs) favors the former. In a realistic application, the MOCOM-UA algorithm is used to calibrate the Soil Moisture Accounting model of the National Weather Service River Forecasting Systems (NWSRFS-SMA). Multi-objective calibration of this model is accomplished using two bi-criterion objective functions, namely the Daily Root Mean Square-Heteroscedastic Maximum Likelihood Estimator (DRMS-HMLE) and rising limb /falling limb (RISE/FALL) objective functions. These two multi-objective calibrations provide some interesting insights into the influence of different objectives in the location of final parameter values, as well as limitations in the structure of the NWSRFS-SMA model.
    • Rainfall estimation from satellite infrared imagery using artificial neural networks

      Hsu, Kuo-Lin; Sorooshian, Soroosh; Gao, Xiaogang; Gupta, Hoshin Vijai; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1997)
      Infrared (IR) imagery collected by geostationary satellites provides useful information about the dirunal evolution of cloud systems. These IR images can be analyzed to indicate the location of clouds as well as the pattern of cloud top temperatures (Tbs). During the past several decades, a number of different approaches for estimation of rainfall rate (RR) from Tb have been explored and concluded that the Tb-RR relationship is (1) highly nonlinear, and (2) seasonally and regionally dependent. Therefore, to properly model the relationship, the model must be able to: (1) detect and identify a non-linear mapping of the Tb-RR relationship; (2) Incorporate information about various cloud properties extracted from IR image; (3) Use feedback obtained from RR observations to adaptively adjust to seasonal and regional variations; and (4) Effectively and efficiently process large amounts of satellite image data in real -time. In this study, a kind of artificial neural network (ANN), called Modified Counter Propagation Network (MCPN), that incorporates these features, has been developed. The model was calibrated using the data around the Japanese Islands provided by the Global Precipitation Climatology Project (GPCP) First Algorithm Intercomparison Project (AIP-I). Validation results over the Japanese Islands and Florida peninsula show that by providing limited ground-truth observation, the MCPN model is effective in monthly and hourly rainfall estimation. Comparison of results from MCPN model and GOES Precipitation Index (GPI) approach is also provided in the study.

      Hsu, Kuo-Lin; Gupta, Hoshin Vijai; Sorooshian, Soroosh; Department of Hydrology & Water Resources, The University of Arizona (Department of Hydrology and Water Resources, University of Arizona (Tucson, AZ), 1996-03)
      A new algorithm is proposed for the identification of three-layer feedforward artificial neural networks. The algorithm, entitled LLSSIM, partitions the weight space into two major groups: the input- hidden and hidden -output weights. The input- hidden weights are trained using a multi -start SIMPLEX algorithm and the hidden -output weights are identified using a conditional linear- least- square estimation approach. Architectural design is accomplished by progressive addition of nodes to the hidden layer. The LLSSIM approach provides globally superior weight estimates with fewer function evaluations than the conventional back propagation (BPA) and adaptive back propagation (ABPA) strategies. Monte -carlo testing on the XOR problem, two function approximation problems, and a rainfall- runoff modeling problem show LLSSIM to be more effective, efficient and stable than BPA and ABPA.