A framework for benchmarking land models

[摘要] Land models, which have been developed by the modeling community in the pastfew decades to predict future states of ecosystems and climate, have to becritically evaluated for their performance skills of simulating ecosystemresponses and feedback to climate change. Benchmarking is an emergingprocedure to measure performance of models against a set of definedstandards. This paper proposes a benchmarking framework for evaluation ofland model performances and, meanwhile, highlights major challenges at thisinfant stage of benchmark analysis. The framework includes (1) targetedaspects of model performance to be evaluated, (2) a set of benchmarks asdefined references to test model performance, (3) metrics to measure andcompare performance skills among models so as to identify model strengths anddeficiencies, and (4) model improvement. Land models are required to simulateexchange of water, energy, carbon and sometimes other trace gases between theatmosphere and land surface, and should be evaluated for their simulations ofbiophysical processes, biogeochemical cycles, and vegetation dynamics inresponse to climate change across broad temporal and spatial scales. Thus,one major challenge is to select and define a limited number of benchmarks toeffectively evaluate land model performance. The second challenge is todevelop metrics of measuring mismatches between models and benchmarks. Themetrics may include (1) a priori thresholds of acceptable modelperformance and (2) a scoring system to combine data–model mismatches forvarious processes at different temporal and spatial scales. The benchmarkanalyses should identify clues of weak model performance to guide futuredevelopment, thus enabling improved predictions of future states ofecosystems and climate. The near-future research effort should be ondevelopment of a set of widely acceptable benchmarks that can be used toobjectively, effectively, and reliably evaluate fundamental properties ofland models to improve their prediction performance skills.

[学科分类] 地球化学与岩石