Theses and Dissertations
ORCID
https://orcid.org/0009-0006-6460-3048
Advisor
Ramirez Avila, John J.
Committee Member
Tarso, Paulo
Committee Member
Swan, J. Edward
Committee Member
Gonzalez, Carlos
Committee Member
Khalid, Mohammad
Date of Degree
5-15-2026
Original embargo terms
Embargo 2 years
Document Type
Graduate Thesis - Open Access
Major
Civil Engineering
Degree Name
Master of Science (M.S.)
College
James Worth Bagley College of Engineering
Department
Richard A. Rula School of Civil and Environmental Engineering
Abstract
The Soil Conservation Service Curve Number (SCS-CN) method underpins U.S. runoff estimation, yet its fixed initial-abstraction ratio 0.20 often overstates flows. Analyses spanning 1,358 watersheds show initial abstraction ratio can vary, with a value of 0.05 better matching observed hydrographs, especially for minor storms. Such conservatism inflates design runoff and drives expensive oversizing of drainage infrastructure. To modernize CN calibration, this thesis unites three investigations. First, regional linear and exponential initial abstraction ratio–CN conversion equations between 0.2 and 0.05 are evaluated to improve translation accuracy. Second, least-squares error and asymptotic calibrations are benchmarked with RSE, NSE, KGE, and PBIAS, and a composite metric, OF_KGE, is proposed to balance bias, variability, and timing while resisting outliers. Third, the optimal pairing—LSE-unsorted calibration with OF_KGE—is applied nationwide, generating a precision-λ map that exposes clear geographic gradients. The resulting data-driven framework replaces one-size-fits-all conservatism with statistically robust, location-specific guidance that enhances runoff prediction and reduces infrastructure costs.
Recommended Citation
Ortiz Garcia, Jesus Rafael Sr, "Advancing curve number hydrology: A CONUS‑Scale framework for initial abstraction ratio calibration, conversion, and precision mapping" (2026). Theses and Dissertations. 6956.
https://scholarsjunction.msstate.edu/td/6956
