ADCME

The ARM Data-Oriented Metrics and Diagnostics Package (ARM-Diags) is a Python-based evaluation tool developed by the ARM Infrastructure Team to facilitate the use of long-term, high-frequency ARM measurements for climate model assessment. The package provides a fully automated framework for computing climatological statistics and generating diagnostic visualizations (Zhang et al., 2020). Users can directly compare their model simulations with ARM observations and assess performance relative to CMIP5 and CMIP6 model ensembles (Zhang et al., 2018; Zheng et al., 2023).

The package integrates three categories of datasets:

1. ARM observational data: Long-term climatologies from SGP, NSA, TWP, ENA, and MAO sites, including surface measurements, radiation, clouds, precipitation, and atmospheric profiles

2. CMIP5/6 model simulations: Multi-model datasets enabling intercomparison and ensemble mean benchmarking

3. Test model data: Sample datasets for package validation and user testing

ARM-Diags is open-source and publicly available at https://github.com/ARM-DOE/arm-gcm-diagnostics.

Version 4.1 enhancements include newly developed land-atmosphere coupling metrics at the SGP site. These process-oriented diagnostics evaluate the coupling strength between land surface conditions and atmospheric boundary layer development through two-legged metrics (assessing terrestrial and atmospheric components separately) and diurnal amplitude metrics (examining propagation of diurnal variations through the surface-atmosphere system).

• Sensible heat flux