Skip to content

Latest commit

 

History

History
35 lines (24 loc) · 2.81 KB

front-matter.md

File metadata and controls

35 lines (24 loc) · 2.81 KB

Report Documentation

Document control

Field Value
Title RMC.BestFit 2.0 Technical Reference Manual
Report number Pending assignment
Version External peer-review draft, prepared for the v2.0.1 release
Date 27 September 2026
Prepared for U.S. Army Corps of Engineers, Risk Management Center
Software checkpoint RMC.BestFit 2.0.1, commit 9427bd27c1ed7d757fde5a37ce24b90a06f3c9ed
Numerical source checkpoint RMC.Numerics commit 7e8e8d1c5f26e045a35ec9fc09367de95ed05b02
Declared Numerics dependency RMC.Numerics 2.2.0
Review status Draft for external peer review - not for public release

Authors

  • C. Haden Smith, U.S. Army Corps of Engineers, Risk Management Center, Lakewood, Colorado, USA.
  • Woodrow L. Fields, U.S. Army Corps of Engineers, Risk Management Center, Lakewood, Colorado, USA.
  • Brian Skahill, Fariborz Maseeh Department of Mathematics and Statistics, Portland State University, Portland, Oregon, USA.

Abstract

RMC.BestFit is a Bayesian-first statistical analysis framework for flood-frequency and related water-resources applications. This manual defines the models, likelihoods, priors, estimation methods, uncertainty calculations, diagnostic measures, parameterizations, numerical conventions, and public scientific interfaces implemented by version 2.0. It follows the application order: time-series data; input data and mixed observations; distribution fitting; univariate and Bulletin 17C analyses; point-process, competing-risk, mixture, and composite models; bivariate copulas and coincident frequency; rating curves; AR, MA, ARIMA, and ARIMAX models; and spatial extremes. Each chapter distinguishes mathematical theory from implemented conventions, states assumptions and limitations, and points to independent verification evidence. The implementation and dependency checkpoints above define the software configuration reviewed by this draft.

Review and release notice

This document is a technical draft supplied for independent scientific and engineering review. A report number, named reviewers, approval signatures, and final distribution statement will be assigned after comment disposition. The draft does not authorize changes to established algorithms, priors, samplers, seeds, tolerances, likelihoods, or reference-result contracts.

Revision record

This edition describes the source checkpoints above. It is a finished draft for external review; scientific approval and public release remain separate decisions. Validation built against the published RMC.Numerics 2.2.0 package, which was built from the stated Numerics commit. The source checkpoint identifies the reviewed implementation, while the retained PDF checksum is recorded in the artifact checksum inventory.