Research Catalog: CGD's Climate Analysis

	Figure 1. The hydrological cycle. Estimates of the main water reservoirs, given in plain font in 103 km3, and the flow of moisture through the system, given in slant font in103 km3/yr, equivalent to Exagrams (1018 g) per year. (Trenberth et al. 2006a). High resolution figure.
	Figure 2. Annual-mean frequency (% of time, left column) and intensity (mm/day, right column) of daily precipitation (>1 mm/day) events from TRMM satellite observations (top panels, 3B42 data set, 1998-2003 mean) and four different coupled models (1991-2000 mean). From Dai 2006a. Note the underestimates of intensity and overestimates of frequency of precipitation in the models.. High resolution figure.

Large-scale moisture in the climate system and models

Global energy and water cycle components contain considerable uncertainties but global constraints can be used to refine estimates over land, deduce ocean heat and water transports as residuals, and establish errors. CAS is updating and extending previous analyses of the energy and moisture budgets in light of the new CERES top-of-atmosphere (TOA) radiation and the ECMWF ERA-40 atmospheric data and interacting to produce new NASA reanalyses (MERRA), as well as follow through with more comprehensive assessments. Over land precipitation P?Evaporation E is computed from the atmospheric moisture budget and is a key input into land surface water storage and runoff.

Continuing progress from last year has occurred especially in global scale water cycle analysis. A key part of this was the production and evaluation of the ERA-40 reanalysis and the water vapor and precipitable water, in particular (Uppala et al. 2005; Trenberth et al. 2005a; Trenberth and Smith 2005). An analysis of the mass of the atmosphere and how well it is conserved was completed, most recently with ERA-40 reanalysis data (Trenberth and Smith 2005). This revealed problems over the southern oceans in pre-satellite days. Using the ERA-40 for 1958 to 2001, adjusted for bias over the southern oceans prior to 1979, an analysis was made of global patterns of monthly mean anomalies of atmospheric mass, which is approximately conserved globally (Trenberth et al. 2005a). The resulting dataset is available from NCAR. This work was featured in an article in Science magazine by R. Kerr (22 October 2004 p 599-600) and is prominent in the IPCC AR4 chapter 3 assessment (Trenberth et al. 2007).

Analyses and evaluations have been performed of global datasets on surface and column-integrated water vapor (precipitable water) (Dai 2006b, Trenberth et al. 2005a). Precipitable water (PW) from both NCEP reanalyses, NVAP, ERA-40, and SSM/I was analyzed and problems were found with them all, with only SSM/I providing reliable trends over the ocean. These works are having considerable impact, for instance, in IPCC AR4 and a figure is featured in AR4. Joint work with EOL has created a global data set of 2-hourly PW using path delay data from ground-based Global Positioning System (GPS) receivers around the world (Wang et al. 2005; Wang et al. 2006). Among other applications, this reliable PW data set will be used to quantify biases in rawinsonde and satellite PW data and to study the diurnal variations in PW.

A study on clouds (Dai et al. 2006) that is highly relevant to IPCC discusses shortcomings in automated surface observation systems (ASOS) that were widely introduced to replace manned weather stations around the middle 1990s over North America and other parts of the world. However, a network of 124 U.S. military weather stations with continuous human observations provides useful information of total cloud cover averaged over the contiguous United States, and suggests an increasing trend (~1.4% of the sky covered per decade) in U.S. total cloud cover from 1976 to 2004, with increases over most of the country except the Northwest. Nevertheless, inadequacies exist in surface observations of global cloud amounts and types, especially over the oceans and over Canada and the United States since the middle 1990s. The problem is compounded by inhomogeneities in satellite cloud data.

Components of the hydrological cycle studied include precipitation (amount, frequency, intensity, type), evapotranspiration (evaporation plus transpiration from plants) (Qian et al. 2006), soil moisture, runoff, streamflow and river discharge into the oceans (Su et al. 2006), atmospheric moisture flows and divergence (Trenberth et al. 2006a), and atmospheric moisture storage (Trenberth and Smith 2005). The historical records and model simulations were analyzed to examine any changes associated with global warming in the water cycle, such as potential drying over land. Related issues are the co-variability of temperature and precipitation (Trenberth and Shea 2005), and forcings of the hydrological cycle, such as solar radiation (Qian et al. 2006). A detailed study of the hydro-meteorology of the Mississippi river basin has been completed (Qian et al. 2007) which utilizes both the energy constraints and water budget constraints to put together a physically consistent picture of trends in the region. Trends from 1948 to 2004 are highly significant and reasonably well depicted by linear trends. Increases in cloud and precipitation have reduced the solar energy available for evaporation (“dimming”) and thus potential evaporation (or pan evaporation), but increased soil moisture has increased evapotranspiration as the actual evaporation has risen to become closer to the potential amount, even as runoff and streamflow have also increased. Diminished solar radiation is offset by reduced outgoing longwave radiation and the increased surface wetness has led to an increase in latent at the expense of a decrease in sensible surface heat flux. This has also been followed up with a new book chapter on monsoons (Trenberth et al. 2005b).

A new estimate has been made of the global hydrological cycle for long-term annual means that includes estimates of the main reservoirs of water as well as the flows of water among them (Trenberth et al. 2006a), see Fig. 1. In addition, the mean annual cycle of the atmospheric hydrological cycle based on 1979 to 2000 data includes monthly estimates of P, evapotranspiration E, atmospheric moisture convergence over land, and changes in atmospheric storage, for the major continental land masses, zonal means over land, hemispheric land means and global land means. The evapotranspiration was computed from the Community Land Model run with realistic atmospheric forcings, including precipitation constrained by observations for monthly means but with high frequency information taken from atmospheric reanalyses. Results for P-E have been contrasted with those from atmospheric moisture budgets based on ERA-40 reanalyses. The latter show physically unrealistic results, because evaporation often exceeds precipitation over land especially in the tropics and subtropics.

Analysis and evaluation of the various components of the global water cycle in climate models includes details on precipitation (Fig. 4) and moist convection (Dai 2006a, Rasch et al. 2006; Sun et al. 2006a,b), atmospheric moisture transport and land surface water fluxes and storage (soil moisture, evaporation, runoff and stream-flow, etc.).

References

Additional Information

Print version (.pdf)

CGD Sectional Narratives