Claude Skills by lzwei196

- **Package version**: 1.0.0 - **Model**: SNOWPACK 3.7.1 - **Domain**: Cryosphere (snow/land-surface) - **Created**: 2026-03-26 - **Updated**: 2026-04-30 (49-site SNOTEL batch validation) - **Tools**: 5 - **Skill documents**: 6 - **Diagnostic triplets**: 28 - **Validation status**: production_validated (western US SNOTEL network) ---

**SUMMA** (Structure for Unifying Multiple Modeling Alternatives) is a flexible multi-physics hydrologic modeling framework developed by NCAR (Clark et al., 2015a,b). Unlike traditional hydrologic models that hardcode one set of process representations, SUMMA lets you **choose** which physics to use for each process via "model decisions" -- then systematically compare alternatives. This knowledge infrastructure enables autonomous AI operation of SUMMA. **Model**: SUMMA (CH-Earth/summa), Fortr...

> **Stage ID**: s1_domain_setup > **Pipeline order**: 1 of 7 > **Depends on**: none

> **Stage ID**: s2_forcing_prep > **Pipeline order**: 2 of 7 > **Depends on**: s1_domain_setup

> **Stage ID**: s3_decisions > **Pipeline order**: 3 of 7 > **Depends on**: none (can run in parallel with Stage 1)

> **Stage ID**: s4_parameters > **Pipeline order**: 4 of 7 > **Depends on**: s1_domain_setup, s3_decisions

> **Stage ID**: s5_initial_conditions > **Pipeline order**: 5 of 7 > **Depends on**: s1_domain_setup, s4_parameters

> **Stage ID**: s6_execution > **Pipeline order**: 6 of 7 > **Depends on**: s1_domain_setup, s2_forcing_prep, s3_decisions, s4_parameters, s5_initial_conditions

> **Stage ID**: s7_physics_comparison > **Pipeline order**: 7 of 7 > **Depends on**: s6_execution

SUMMA has **no internal channel routing**. `averageRoutedRunoff` is the sub-grid gamma time-delay histogram, not channel discharge. Gauged-streamflow validation therefore requires this stage. See `dag.yaml` hazards: `averageRoutedRunoff_scored_as_channel_discharge`, `mizuroute_hybas_id_int32_overflow_and_unit_double_conversion`, `lumped_discretisation_cannot_be_routed`.

**Package**: `pyswan-ocean` v1.0.0 **Model**: PySWaN — Python I/O and spectral analysis toolbox for SWAN (Simulating WAves Nearshore) **Source**: [github.com/openearth/swan](https://github.com/openearth/swan) **Author**: Gerben de Boer (TU Delft / Van Oord) **Last updated**: 2026-03-26 **Stats**: 4 tools | 5 skill documents | 20 diagnostic triplets | ~1,200 lines of validated Python **Validation status**: `tested` (round-trip JONSWAP spectrum generation, write, read, Hm0 verification) ---

**Package**: `hydrocraft-swap-soil` v1.0.0 **Model**: SWAP v4.2.0 **Developer**: Wageningen University and Research (WUR), The Netherlands **Last updated**: 2026-03-28 **Stats**: 6 tools | 6 skill documents | 23 diagnostic triplets | ~1,600 lines of validated Python **Validation status**: `analytic` (Hupselbrook, NL, 2002-2004) + `real-case` (FLUXNET US-Ne1, 2001-2013) ---

Define the simulation site, period, forcing source, crop schedule, and output options. This stage establishes all parameters needed to assemble a complete SWAP .swp file.

Convert external meteorological data (CMFD, MSWX, or other sources) to the SWAP .met file format with correct units and structure.

Convert soil texture/property data from HWSD or other databases to SWAP Mualem-van Genuchten (MvG) hydraulic parameters and profile discretization.

Run the SWAP binary with validated input files and verify successful completion through output file checks and water balance closure.

Extract and structure SWAP output files into standardized CSV format for analysis, visualization, and comparison with observations.

Generate publication-quality plots of SWAP simulation results and compare against observations for model validation.

python tools/s1/delineate_watershed.py dem_clip.tif <lat> <lon> delin/ 25 0.01 python tools/s1/build_channel_topology.py \ --subbasin_raster delin/subbasins.tif --dem_path dem_clip.tif \ --flow_dir delin/flow_direction.tif --flow_acc delin/flow_accumulation.tif \ --streams_shp delin/streams.shp \ --repair_spurious_outlets \ --output delin/channel_topology.json python tools/s2/generate_hru_from_global.py --basin_shp delin/watershed.shp \ --dem_path dem_clip.tif --output_dir TxtInOut --basin_na...

> **Stage ID**: s1_watershed_delineation > **Pipeline order**: 1 of 9 > **Depends on**: none

> **Stage ID**: s2_hru_definition > **Pipeline order**: 2 of 9 > **Depends on**: s1_watershed_delineation

> **Stage ID**: s3_weather_preparation > **Pipeline order**: 3 of 9 > **Depends on**: s1_watershed_delineation (for station locations)

> **Stage ID**: s4_soil_database > **Pipeline order**: 4 of 9 > **Depends on**: none (independent — can run in parallel with S1-S3)

> **Stage ID**: s5_landuse_management > **Pipeline order**: 5 of 9 > **Depends on**: s2_hru_definition

> **Stage ID**: s6_calibration_parameters > **Pipeline order**: 6 of 9 > **Depends on**: s2_hru_definition, s4_soil_database

> **Stage ID**: s7_simulation_config > **Pipeline order**: 7 of 9 > **Depends on**: s1_watershed_delineation, s2_hru_definition, s3_weather_preparation, s4_soil_database, s5_landuse_management, s6_calibration_parameters

> **Stage ID**: s8_model_execution > **Pipeline order**: 8 of 9 > **Depends on**: s7_simulation_config

> **Stage ID**: s9_output_parsing > **Pipeline order**: 9 of 9 > **Depends on**: s8_model_execution

**Model**: EPA SWMM 5.2 (Storm Water Management Model) **Developer**: US Environmental Protection Agency / Computational Hydraulics International (CHI) **Engine**: C library (open source), accessed via pyswmm (Python) or swmm-toolkit **Domain**: Urban stormwater drainage, pipe network hydraulics, green infrastructure (LID), urban flood modeling **Repository**: https://github.com/USEPA/SWMM (C engine), https://github.com/pyswmm/pyswmm (Python API) **Documentation**: https://www.epa.gov/water-r...

> **Stage ID**: s1_subcatchment_delineation > **Pipeline order**: 1 of 7 > **Depends on**: none

> **Stage ID**: s2_drainage_network > **Pipeline order**: 2 of 7 > **Depends on**: none (can run in parallel with S1 and S3)

> **Stage ID**: s3_rainfall_forcing > **Pipeline order**: 3 of 7 > **Depends on**: none (can run in parallel with S1 and S2)

> **Stage ID**: s4_lid_setup > **Pipeline order**: 4 of 7 > **Depends on**: s1_subcatchment_delineation

> **Stage ID**: s5_model_assembly > **Pipeline order**: 5 of 7 > **Depends on**: s1_subcatchment_delineation, s2_drainage_network, s3_rainfall_forcing (optionally s4_lid_setup)

> **Stage ID**: s6_execution > **Pipeline order**: 6 of 7 > **Depends on**: s5_model_assembly

> **Stage ID**: s7_model_coupling > **Pipeline order**: 7 of 7 > **Depends on**: s6_execution (and VIC/CaMa-Flood simulations from HydroCraft)

**Package**: simfire-wildfire v2.0.1 **Target Model**: SimFire 2.0.1 (Rothermel Surface Fire Spread) **Domain**: Wildfire simulation, fire behavior prediction, reinforcement learning **Authors**: MITRE Fireline (Welsh, Dotter, Doyle, Gandikota, Kempis, Schambach, Tapley, Threet) **Last updated**: 2026-08-09 **Stats**: 6 tools, 5 skill docs, 25 diagnostic triplets, ~10,600 lines Python source ---

| Field | Value | |----------------|-------------------------------------------------------------| | **Package** | SimPEG KI v1.0 | | **Model** | SimPEG (Simulation and Parameter Estimation in Geophysics) | | **Domain** | Geophysics (forward modeling and inversion) | | **Language** | Python (>=3.11) | | **Authors...

**Package**: superflexpy-ki v1.0.0 **Model**: SuperflexPy v1.3.2 — Flexible Conceptual Hydrological Modelling Framework **Authors**: Marco Dal Molin, Fabrizio Fenicia, Dmitri Kavetski **Created by**: Hydrocraft / auto_dissect **Last updated**: 2026-03-25 **Stats**: 4 tools | 5 skill docs | 21 diagnostic triplets | ~2800 lines ---

| Field | Value | |--------------------|-------------------------------------------------------------| | Package | `telemac-mascaret-ki` | | Model | TELEMAC-MASCARET v9.1.0 | | Domain | Free-surface flow, wave propagation, sediment transport | | Language | Fortran 90 / Python 3 ...

**Package**: `hydrocraft-topmodel` v1.0.0 **Model**: TOPMODEL BMI (Beven & Kirkby 1979, NOAA-OWP C implementation) **Created by**: Jianyun Zhang Research Group, Hohai University **Last updated**: 2026-03-28 **Stats**: 5 tools | 6 skill documents | 18 diagnostic triplets | ~2,100 lines of validated Python **Validation status**: see `knowledge_infrastructure.yaml` ---

**Package**: `hydrocraft-trigrs-landslide` v1.0.0 **Model**: TRIGRS v2.1.00c (serial) + MPI parallel version **Source**: USGS (Rex L. Baum & Massimiliano Alvioli) **Last updated**: 2026-03-26 **Stats**: 5 tools | 6 skill documents | 18 diagnostic triplets | ~2,000 lines of validated Python **Validation status**: `tutorial_validated` (USGS tutorial dataset, Srivastava-Yeh column test) ---

| Field | Value | |-------------------|--------------------------------------------| | Package | topoflow-ki | | Version | 1.0.0 | | Target Model | TopoFlow 3.6 | | Domain | Spatially-distributed hydrology | | Author | Scott D. Peckham (model); KI auto-dissect | | Created | 2026-0...

**Package**: hydrocraft-velma v1.0.0 **Model**: VELMA (Visualizing Ecosystem Land Management Assessments) **Domain**: Multi-layer ecohydrological soil water balance **Language**: Java (original, USEPA); Python (analytic reimplementation) **License**: Public domain (USEPA) **Created**: 2026-03-30 **Validation**: Bengbu station (51080), Huai River Basin, China (1980--1990) | Metric | Value | |--------|-------| | Tools | 4 | | Pipeline stages | 7 | | Diagnostic triplets | 14 | | Validation basin...

``` Expected runtime depends on basin size and period. The Harbin reference noted below used 866 cells; water-balance mode costs about 95 s per simulated year for that case. ---

**Package**: hydrocraft-wasp v1.0.0 **Model**: WASP (Water Analysis Simulation Program) -- Analytic Reimplementation **Domain**: Lake and reservoir water quality modeling **Language**: Python (analytic reimplementation of WASP core physics) **License**: Public domain (EPA) **Created**: 2026-03-30 **Validation**: Lake Erie Central Basin, WQP observations (Temp R=0.885, DO Cal R=0.816) | Metric | Value | |--------|-------| | Tools | 4 | | Pipeline stages | 6 | | Diagnostic triplets | 26 | | Val...

**Model**: WOFOST 7.2 (WOrld FOod STudies) via PCSE 6.0 (Python Crop Simulation Environment) **Domain**: Crop growth simulation / food production modelling **Created by**: Zhang Jianyun Research Group, Hohai University ---

> **Stage ID**: s1_crop_params > **Pipeline order**: 1 of 8 > **Depends on**: none

> **Stage ID**: s2_soil_params > **Pipeline order**: 2 of 8 > **Depends on**: none

> **Stage ID**: s3_weather_prep > **Pipeline order**: 3 of 8 > **Depends on**: none