Capabilities by Task

Choose an operation from the inputs you already have and the output you need. Each row links to the guide, API contract, and a concrete example. For exact signatures and return fields, use the linked API reference; for model-specific settings and retained evidence, read the example's configuration and results.

Task	Inputs → outputs	Guide and API	Examples
Discover and inspect projects	Project files → plan, geometry, boundary, and result inventories with resolved paths	Initialization · RasPrj / RasPlan	101: project inventory, 104: plan parameters
Execute plans and inspect completion	Project, plan, selected runtime → result object, native artifacts, messages, and completion evidence	Execution and return contracts · RasCmdr · Remote execution	110: one plan, 113: parallel plans
Read 1D results	Steady or unsteady HDF → profile-indexed WSE or cross-section time series	HDF extraction · Steady results · HDF API	400: unsteady 1D, 401: steady profiles
Automate 2D line plots and hydrographs	Lines and saved results, or reference locations authored before a new run → station profiles, reference time series and batch figures	Choose an extraction route · Profiles and reference API	314: author and compute references, 413: line flow, 416: velocity profiles
Query 2D results	Mesh geometry and results → cell, face, point, or line quantities with spatial/time coordinates	HDF extraction and result shapes · HdfResultsMesh	410: mesh extraction, 412: faces, 413: line flow
Author geometry and structures	Working geometry, terrain, or land-cover inputs → edited model plus readback and native-processing evidence	Geometry operations · Geometry API	212: Manning layers, 216: bridges, 227: connection culverts
Author steady flows	Profile discharges and typed boundaries → steady-flow files and linked plans	Steady-flow guide · RasSteady	224: steady-flow authoring
Run containers or a cluster campaign	Complete working models or solver-ready artifacts → receipts, logs, final HDFs, and scheduler evidence	Choose a backend · Docker Hub catalog · Slurm + Apptainer · Portable Slurm	512: Docker preprocessing and native computation
Prepare precipitation and boundaries	Observed, design, or forecast data → time/units conversion, hyetographs, DSS, or model forcing	Boundary conditions · Gridded precipitation · DSS API	313: HMS/DSS matching, 720: storm methods, 923: coastal stages
Calibrate and explore uncertainty	Model, declared targets/observations, parameter design → objective values, candidates, and ensemble diagnostics	Workflow patterns · Calibration and uncertainty APIs · Gauge data	220: calibration, 117: uncertainty, 922: delivered-model/gauge comparison
Acquire and inspect model sources	Catalog or archive → identified projects, provenance, inventories, and delivered-result inspection	eBFE delivery guide · Model-source API	950: eBFE organization, 958: source catalogs
Conflate or extract a submodel	Network edge or proposed domain plus source models → coverage/selection plan, assembled project or prepared geometry, seam/flux review	Hydrofabric · 1D breakout · 2D breakout	235: 1D extraction, 236: multi-model assembly, 959: 2D breakout and boundary diagnostics
Generate GIS and result products	Compatible geometry/results or aligned depth rasters → spatial exports, product manifests, checksums, and comparison maps	Cloud-native export · HdfResultsProducts · BenefitArea	417: retained-result products, 612: benefit/impact maps, 962: COG export
Screen model diagnostics	Geometry and computed outputs → scoped messages, tables, and review reports	Quality assurance and RasCheck calls	800: steady-model diagnostics, 801: structures

Floodway authoring and checking remain experimental. The Floodway Check contract records parser, dataset-access, coverage, and qualification limitations. The steady-flow authoring example above does not qualify encroachment trials. There is currently no qualified floodway notebook route in this catalog.

Choose the Runtime for the Operation

Operation	Requirement to establish before running
Inspect text, HDF, or retained results	Python dependencies and the required stored datasets; reading results does not require a new solve
Edit text geometry, plans, or boundaries	A writable working copy; readback confirms stored inputs, while dependent preprocessing/results may need regeneration
Native geometry or sidecar authoring	The API's specified HEC-RAS/RASMapper version and platform; validate associations before interpreting final cell/face values
Execute a plan	A compatible installed solver or configured remote runtime, the exact project/plan, and the applicable runtime preflight
Use COM or GUI workflows	Windows, the selected registered/native HEC-RAS version, and an interactive desktop where required
Export products or compare observations	Compatible source quantity, coordinates, units, vertical/time reference, and declared output/metric settings

Consult installation requirements, the gridded-precipitation route guidance, and the operation's API restrictions. A library import or documentation build does not qualify every solver/platform combination. The release notes distinguish installed releases from later documentation and development changes.

Interpret the Evidence

Inventory means files were identified; readback means stored fields were read; preprocessing means native inputs were prepared; completion means the execution contract was checked. Numerical diagnostics, observation comparisons, and engineering acceptance answer additional questions. Keep those claims separate when using a result for a new model or study purpose.

For example, a signed profile-line flux differs from summed absolute face flows; a maximum differs from an instantaneous sample; missing output differs from zero; and a benefit-area category does not describe every physical change. The linked contracts specify these distinctions where they apply.

The complete example catalog provides task summaries and selected expandable input/output/runtime/evidence contracts. Its machine-readable JSON exposes the same curated fields from examples/notebooks.yml. Missing optional metadata means the field has not been curated. Saved output and recorded execution time remain separate from workflow qualification.