Noah_MP

Runs and calibrates Noah-MP land surface simulations via the HRLDAS offline driver.

155|6|Updated Mar 31, 2026
One-click install
npx skills add https://github.com/lzwei196/KISS-Knowledge-Infrastructure-for-Scientific-Simulation --skill noah-mp-lzwei196
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Noah_MP
Source: https://github.com/lzwei196/KISS-Knowledge-Infrastructure-for-Scientific-Simulation/tree/main/models/Noah_MP
Command: npx skills add https://github.com/lzwei196/KISS-Knowledge-Infrastructure-for-Scientific-Simulation --skill noah-mp-lzwei196

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires netCDF4, numpy, pandas, xarray, matplotlib, scipy, PyYAML, and includes scripts (resource) and references (resource) components.

What problem does it solve? Running the Noah-MP land surface model offline requires deep operational knowledge: exact forcing units (Kelvin, mm/s, Pa, specific humidity), HRLDAS setup NetCDF structure, namelist physics options, and recovery from silent failures. This Skill packages that expertise so an AI agent can prepare inputs, execute the real HRLDAS binary, parse outputs, and calibrate parameters without hand-coded approximations. ## Core Features & Use Cases - End-to-end HRLDAS pipeline: Six validated Python tools cover setup-file generation, soil conversion (HWSD/SoilGrids), forcing conversion (CMFD/MSWX/ERA5/FLUXNET to LDASIN), model execution, and LDASOUT parsing. - Diagnostic recovery: 29 symptom-diagnosis-remedy triplets in diagnostics/triplets.yaml map known failure modes (unit errors, rank mismatches, irrigation misconfiguration) to validated fixes. - Site-level calibration: A commissioning-tested calibration contract (calibration.yaml) injects 14 parameters into NoahmpTable.TBL and scores latent heat against FLUXNET observations with NSE, including blocked temporal holdout. - Use Case: Drive a FLUXNET tower site run (e.g., US-Ne1 irrigated maize) from raw tower meteorology to a validated evapotranspiration time series, with crop and irrigation options correctly wired. ## Quick Start Run python preflight_check.py in this directory, then ask the agent to convert FLUXNET forcing for your site and execute the Noah-MP HRLDAS run with the provided tools.

Frequently Asked Questions about Noah_MP

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run Noah-MP offline with HRLDAS?

Build the HRLDAS driver from source with gfortran and NetCDF-Fortran, then use the provided tools: build_hrldas_setup.py for the domain file, convert_forcing_to_noahmp.py for LDASIN forcing, and run_noahmp.py to execute the binary. Run preflight_check.py first to verify the environment.

How do I convert FLUXNET tower data to Noah-MP forcing?

Use convert_forcing_to_noahmp.py with --source fluxnet, which reads FULLSET gap-filled tower meteorology and writes hourly LDASIN NetCDF files. The --utc_offset argument is mandatory because FLUXNET timestamps are local standard time while HRLDAS reads them as UTC.

What units does Noah-MP forcing data require?

Noah-MP requires T2D in Kelvin, RAINRATE in mm/s, Q2D as specific humidity in kg/kg, PSFC in Pascals, and SWDOWN/LWDOWN in W/m2. Passing Celsius, mm/hr, or relative humidity causes silent failures or crashes, each covered by a diagnostic triplet.

Why does my Noah-MP run produce extreme runoff or NaN values?

Check diagnostics/triplets.yaml first: NaN in vapor pressure usually means Celsius instead of Kelvin (dt_001), and extreme runoff means precipitation in mm/hr instead of mm/s (dt_002). Negative interpolated shortwave radiation also causes energy-balance errors (dt_004).

Can Noah-MP simulate irrigation at cropland sites?

Yes, with CROP_OPTION=1 and IRRIGATION_OPTION set, but the land use must be a MODIS cropland class and irrigation fractions must be in a separate AGDATA file. build_hrldas_setup.py writes both files and validates ranks and activation thresholds.

What are the limitations of offline Noah-MP column runs?

The offline HRLDAS mode does not simulate the atmosphere, so forcing must be supplied, and it performs no lateral river routing. Column runoff is not streamflow; comparing to gauged discharge requires an external routing model such as WRF-Hydro.