1. IPCC Model Uncertainty in Representing the
Global Water Cycle of the late 20th Century Climate
Duane Waliser, JPL
Water Cycle quantities are ordered left -> right in increasing uncertainty.
Fluxes (red) show better agreement than reservoirs (blue).
Vapor ( ) and Liquid ( ) show better agreement than Frozen ( ) water.
Taking the above into consideration, it can be seen that quantities that have robust observational constraints, in many cases from satellite, tend to exhibit less model uncertainty.
The most uncertain water cycle components (e.g., snow, ice, cloud mass) represent key climate feedbacks.
Complete Caption (top) Model-to-model agreement in globally-averaged, annual mean values of important hydrological quantities from the period of the 20th century GCM simulations contributed to the IPCC 4th Assessment Report (20c3m scenario). Normalized percent differences are relative to the multi-model mean (MMM) values [e.g., 16 models provide soil moisture values, the standard deviation is about +/- 60% of the MMM, with a maximum (minimum) of about +100% (-100%)]. Quantities are ordered in increasing model disagreement using the standard deviation. (bottom) Same, except for expanded y-scale. Horizontal labels consist of the name model variable and the number of model contributions included in the analysis. Font color indicates whether the water cycle component is a flux (red) or reservoir (blue). In addition, model variables are labeled with icons indicating whether the flux or reservoir is associated with vapor (molecule), liquid (drop) and/or ice (snowflake).