A Multi-model Evaluation of the Water Vapor Budget in Atmospheric Rivers

Atmospheric rivers (ARs) are narrow regions of strong horizontal water vapor transport that 27 play important roles in the global water cycle, weather, and hydrology. Motivated by challenges in 28 simulating ARs with state-of-the-art global models, this paper diagnoses model errors with a focus on 29 relative contributions of moisture convergence, evaporation, and precipitation to AR column-integrated 30 water vapor (IWV) budget. 31 Using 20-year simulations by 24 global weather/climate models, budget terms are calculated for 32 four AR sectors: post-frontal, frontal, pre-frontal, and pre-AR, with biases assessed against two 33 reanalysis products. The results indicate that each sector is unique in terms of the dominant water vapor 34 balance, and that the terms exhibiting the largest inter-model spread are the same terms dominating the 35 water vapor balance in each sector. Overall, simulated bulk AR characteristics (e.g., geometry, 36 frequency, intensity) are more sensitive to biases in IVT convergence and IWV tendency than to biases 37 in evaporation and precipitation, although evaporation/precipitation biases do affect key AR bulk 38 characteristics in selected sectors. The large inter-model spread (particularly for precipitation) and, in 39 certain cases, discrepancies between the reanalysis references themselves (particularly for precipitation 40 types), highlight the need for observational efforts that target better constraining AR processes in 41 weather/climate models and reanalyses.