Evaluation of the ACME Land Model using the ILAMB Prototype

Abstract

Assessment of overall model fidelity requires a comprehensive comparison of model results with a wide variety of observational data spanning multiple space and time scales. The International Land Model Benchmarking (ILAMB) activity has developed an open source benchmarking software system that employs a growing collection of laboratory, field, and remote sensing data sets for systematic evaluation of terrestrial biogeochemical and biogeophysical processes. The ACME Project is leveraging the ILAMB metrics and diagnostics prototype system, developed by the Biogeochemistry–Climate Feedbacks Scientific Focus Area (SFA), and extending the system to assess the overall performance of the ACME Land Model (ALM), both coupled and uncoupled, as it evolves over time. Here we show how the performance of the latest version of ALM, run offline and forced with CRU-NCEP reanalysis, compared with the performance of the Community Land Model (CLM) versions CLM4.0-CN (forced with CRU-NCEP), CLM4.5-BGC (forced with CRU-NCEP), and CLM4.5-BGC (forced with GSWP3) on the full suite of metrics currently contained in the ILAMB system. Diagnostics highlighting key model response differences attributable to new model structures implemented in ALM will be presented. Additional proposed metrics important for evaluating new process representations in ALM will also be discussed.