393945 - Exploration of Dynamically Dimensioned Search Algorithm: A Deeper Look of Efficiency/Optimization

Monday, June 4
10:30 AM - 12:00 PM
Location: Greenway IJ

The Dynamically Dimensioned Search algorithm (DDS) is an automatic calibration and optimization technique designed to solve challenging calibration problems for complex watershed simulation models. The primary advantage of DDS is that it outperforms many other optimization techniques in both convergence speed and the ability in searching for parameter sets that satisfy statistical guidelines while requiring only one system parameter (perturbation factor) in the optimization process. A default value of 0.2 is generally used as the perturbation factor, but the sensitivity of perturbation factor to the performance in model calibration is still unknown. In addition, the sampling distribution used in DDS is fixed with the mean value of zero and the variance that equals to one. The impact of model prediction caused by the selection of sampling distribution is still unidentified. The goal of this study is to evaluate the efficiency of DDS by altering sampling ranges and sampling distributions on hydrologic and water quality prediction. In case study, the Soil and Water Assessment Tool (SWAT) was adopted to validate the efficiency of different versions of DDS on a watershed with both flow and water quality data. Results show that the use of various values of perturbation factor can cause variations in convergence speed but not in the performance of finding better solutions. However, the selection of sampling distribution may alter calibration results considerably.

Haw Yen, PhD

Assistant Research Scientist
Texas A&M University

Dr. Yen has interests in watershed modeling; point and nonpoint sources pollution control/simulation; traditional and heuristic optimization techniques; sensitivity & uncertainty analysis; water resources planning & management; reservoir operation rules development & real-time flood control operation; and data management. Dr. Yen has developed the Integrated Parameter Estimation and Uncertainty Analysis Tool (IPEAT) to incorporate uncertainty sources from system parameters, forcing inputs, measured data and model structure. IPEAT has been implemented to large-scale national projects such as the USDA funded Conservation Effects Assessment Project (CEAP), and the EPA funded Hydrologic And Water Quality System (HAWQS). At the Blackland Research and Extension Center, Texas A & M University, he designs and develops watershed simulations to mitigate environmental and ecological impacts for the original and the new version of the Soil and Water Assessment Tool (SWAT and SWAT+) as well as participating in SWAT training seminars and workshops dealing with new technology. In addition, he is supporting the automated systems for the Agricultural Policy/Environmental eXtender (APEX) such as the APEX autocalibration tool (APEX-CUTE). Dr. Yen is a member of the American Society of Civil Engineers (ASCE) and the American Geophysical Union (AGU). Dr. Yen is also actively serving multiple associate editor positions in internationally recognized journals such as Limnology [Springer, ISSN: 1439-8621].


Send Email for Haw Yen


393945 - Exploration of Dynamically Dimensioned Search Algorithm: A Deeper Look of Efficiency/Optimization

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Exploration of Dynamically Dimensioned Search Algorithm: A Deeper Look of Efficiency/Optimization