Abstract

The NCEP CFSv2 ensemble reforecasts initialized with different land surface analyses for the period of 1979–2010 have been conducted to assess the effect of uncertainty in land initial states on surface air temperature prediction. The two observation-based land initial states are adapted from the NCEP CFS Reanalysis (CFSR) and the NASA GLDAS-2 analysis; atmosphere, ocean, and ice initial states are identical for both reforecasts. This identical-twin experiment confirms that the prediction skill of surface air temperature is sensitive to the uncertainty of land initial states, especially in soil moisture and snow cover. There is no distinct characteristic that determines which set of the reforecasts performs better. Rather, the better performer varies with the lead week and location for each season. Estimates of soil moisture between the two land initial states are significantly different with an apparent north–south contrast for almost all seasons, causing predicted surface air temperature discrepancies between the two sets of reforecasts, particularly in regions where the magnitude of initial soil moisture difference lies in the top quintile. In boreal spring, inconsistency of snow cover between the two land initial states also plays a critical role in enhancing the discrepancy of predicted surface air temperature from week 5 to week 8. Our results suggest that a reduction of the uncertainty in land surface properties among the current land surface analyses will be beneficial to improving the prediction skill of surface air temperature on subseasonal time scales. Implications of a multiple land surface analysis ensemble are also discussed.

Abstract

The simplified Simple Biosphere model (SSiB) has been validated using observed meteorological, turbulent flux, and vegetation property data from the Anglo-Brazilian Amazonian Climate Observation Study (ABRACOS) over a forest clearing site. The results show that SSiB is able to simulate the observed fluxes realistically. The differences between the simulated and observed latent and sensible heat fluxes are less than 10 W m⁻². Compared to previous deforestation experiments, the new vegetation dataset produces significantly different latent heat fluxes and surface temperatures in off-line and general circulation model (GCM) simulation. Using the new dataset the GCM simulated surface temperature is about 2 K higher, and the simulated latent heat flux is about 25 W m⁻² lower than that generated using a previous dataset. These differences can be expected to result in substantially different responses in rainfall and atmosphere circulation. The parameters that are most significant in producing such large differences are leaf area index and soil properties. This study again demonstrates that to realistically assess the climatic impact of land surface degradation a realistic specification of the land surface conditions within GCMs is crucial.

Abstract

The terrestrial and oceanic sources of moisture that supply warm-season rainfall to the Mississippi River basin and its subbasins are examined over a 36-yr period (1963–98). Using hourly observed precipitation, National Centers for Environmental Prediction (NCEP) reanalyses at 6-h intervals, and a back-trajectory algorithm, the water falling during observed precipitation events is probabilistically traced to its most recent surface evaporative source, terrestrial or oceanic. Maps of these sources generally show dual maxima, one terrestrial and one oceanic, in spring and a dominance of terrestrial sources in summer. Pentad time series averaged over the 36 years show a late-summer maximum of precipitation recycling in all but the Missouri subbasin. During the 36 years analyzed, 32% of warm-season precipitation in the entire Mississippi basin originated as evaporation within the basin (recycled). About 20% of warm-season precipitation was contributed directly by evaporation from the Gulf of Mexico and Caribbean. The Midwest flood year, 1993, represents a positive outlier in terms of July precipitation supplied to the Upper Mississippi directly by evaporation from the Caribbean. The monthly recycling ratios for warm-season precipitation during the drought year, 1988, represent extreme values in the time series but are not identified as outliers. A positive trend in precipitation recycling in the Upper Mississippi and Missouri subbasins and accompanying decrease in Gulf of Mexico/Caribbean–supplied precipitation to those regions are statistically significant but may reflect changes in the observational data stream assimilated by the NCEP model. Perturbation analysis demonstrates that the source fractions and recycling ratios are somewhat sensitive to systematic errors but not to random errors in the model-derived evapotranspiration (ET), arguably the largest source of uncertainty in the back-trajectory approach. Systematic errors in terrestrial ET on the order of 20% introduce errors of about 0.02 in land source fractions (including recycling ratios) that are themselves on the order of 0.10–0.30.

Abstract

In addition to remote SST forcing, realistic representation of land forcing (i.e., soil moisture) over the United States is critical for a prediction of U.S. severe drought events approximately one season in advance. Using “identical twin” experiments with different land initial conditions (ICs) in the 32-yr (1979–2010) CFSv2 reforecasts (NASA GLDAS-2 reanalysis versus NCEP CFSR), sensitivity and skill of U.S. drought predictions to land ICs are evaluated. Although there is no outstanding performer between the two sets of forecasts with different land ICs, each set shows greater skill in some regions, but their locations vary with forecast lead time and season. The 1999 case study demonstrates that although a pattern of below-normal SSTs in the Pacific in the fall and winter is realistically reproduced in both reforecasts, GLDAS-2 land initial states display a stronger east–west gradient of soil moisture, particularly drier in the eastern United States and more consistent with observations, leading to warmer surface temperature anomalies over the United States. Anomalies lasting for one season are accompanied by more persistent barotropic (warm core) anomalous high pressure over CONUS, which results in better prediction skill of this drought case up to 4 months in advance in the reforecasts with GLDAS-2 land ICs. Therefore, it is essential to minimize the uncertainty of land initial states among the current land analyses for improving U.S. drought prediction on seasonal time scales.

Abstract

The ability of phase 5 of the Coupled Model Intercomparison Project (CMIP5) climate models to simulate the twentieth-century “warming hole” over North America is explored, along with the warming hole’s relationship with natural climate variability. Twenty-first-century warming hole projections are also examined for two future emission scenarios, the 8.5 and 4.5 W m⁻² representative concentration pathways (RCP8.5 and RCP4.5). Simulations from 22 CMIP5 climate models were analyzed, including all their ensemble members, for a total of 192 climate realizations. A nonparametric trend detection method was employed, and an alternative perspective emphasizing trend variability. Observations show multidecadal variability in the sign and magnitude of the trend, where the twentieth-century temperature trend over the eastern United States appears to be associated with low-frequency (multidecadal) variability in the North Atlantic temperatures. Most CMIP5 climate models simulate significantly lower “relative power” in the North Atlantic multidecadal oscillations than observed. Models that have relatively higher skill in simulating the North Atlantic multidecadal oscillation also are more likely to reproduce the warming hole. It was also found that the trend variability envelope simulated by multiple CMIP5 climate models brackets the observed warming hole. Based on the multimodel analysis, it is found that in the twenty-first-century climate simulations the presence or absence of the warming hole depends on future emission scenarios; the RCP8.5 scenario indicates a disappearance of the warming hole, whereas the RCP4.5 scenario shows some chance (10%–20%) of the warming hole’s reappearance in the latter half of the twenty-first century, consistent with CO₂ stabilization.

Abstract

An atmospheric general circulation model (AGCM) is coupled to three different land surface schemes (LSSs), both individually and in combination (i.e., the LSSs receive the same AGCM forcing each time step and the averaged upward surface fluxes are passed back to the AGCM), to study the uncertainty of simulated climatologies and variabilities caused by different LSSs. This tiling of the LSSs is done to study the uncertainty of simulated mean climate and climate variability caused by variations between LSSs. The three LSSs produce significantly different surface fluxes over most of the land, no matter whether they are coupled individually or in combination. Although the three LSSs receive the same atmospheric forcing in the combined experiment, the inter-LSS spread of latent heat flux can be larger or smaller than the individually coupled experiment, depending mostly on the evaporation regime of the schemes in different regions. Differences in precipitation are the main reason for the different latent heat fluxes over semiarid regions, but for sensible heat flux, the atmospheric differences and LSS differences have comparable contributions. The influence of LSS uncertainties on the simulation of surface temperature is strongest in dry seasons, and its influence on daily maximum temperature is stronger than on minimum temperature. Land–atmosphere interaction can dampen the impact of LSS uncertainties on surface temperature in the tropics, but can strengthen their impact in middle to high latitudes. Variations in the persistence of surface heat fluxes exist among the LSSs, which, however, have little impact on the global pattern of precipitation persistence. The results provide guidance to future diagnosis of model uncertainties related to LSSs.

Abstract

The discharge of freshwater into oceans represents a fundamental process in the global climate system, and this flux is taken into account in simulations with general circulation models (GCMs). Moreover, the availability of realistic river routing schemes is a powerful instrument to assess the validity of land surface components, which have been recognized to be crucial for the global climate simulation. In this study, surface and subsurface runoff generated by the 13 land surface schemes (LSSs) participating in the Second Global Soil Wetness Project (GSWP-2) are used as input fields for the Hydrology Discharge (HD) routing model to simulate discharge for 30 of the world’s largest rivers. The simplest land surface models do not provide a good representation of runoff, and routed river flows using these inputs are affected by many biases. On the other hand, HD shows the best simulations when forced by two of the more sophisticated schemes. The multimodel ensemble GSWP-2 generates the best phasing of the annual cycle as well as a good representation of absolute values, although the ensemble mean tends to smooth the peaks. Finally, the intermodel comparison shows the limits and deficiencies of a velocity-constant routing model such as HD, particularly in the phase of mean annual discharge.

The second part of the study assesses the sensitivity of river discharge to the variation of external meteorological forcing. The Center for Ocean–Land–Atmosphere Studies version of the SSiB model is constrained with different meteorological fields and the resulting runoff is used as input for HD. River flow is most sensitive to precipitation variability, but changes in radiative forcing affect discharge as well, presumably because of the interaction with evaporation. Also, this analysis provides an estimate of the sensitivity of river discharge to precipitation variations. A few areas (e.g., central and eastern Asia, the Mediterranean, and much of the United States) show a magnified response of river discharge to a given percentage change in precipitation. Hence, an amplified effect of droughts as indicated by the consensus of climate change predictions may occur in places such as the Mediterranean. Conversely, increasing summer precipitation foreseen in places like southern and eastern Asia may amplify floods in these poor and heavily populated regions. Globally, a 1% fluctuation in precipitation forcing results in an average 2.3% change in discharge. These results can be used for the definition and assessment of new strategies for land use and water management in the near future.

Abstract

The change in predictability of monthly mean temperature in a future climate is quantified based on the Community Climate System Model, version 4. According to this model, the North Atlantic overtakes the El Niño–Southern Oscillation (ENSO) as the dominant area of seasonal predictability by 2095. This change arises partly because ENSO becomes less variable and partly because the ENSO teleconnection pattern expands into the Atlantic. Over land, the largest change in temperature predictability occurs in the tropics and is predominantly due to a decrease in ENSO variability. The southern peninsula of Africa and northeast South America are predicted to experience significant drying in a future climate, which decreases the effective heat capacity and memory, and hence increases variance independently of ENSO changes. Extratropical land areas experience enhanced precipitation in a future climate, which decreases temperature variance by the same mechanism. Finally, the model predicts that surface temperatures near the poles will become more predictable and less variable in a future climate, primarily because melting sea ice exposes the underlying sea surface temperature, which is more predictable owing to its longer time scale. Some of these results, especially the change in ENSO variance, are known to be model dependent. This paper also advances the use of information theory to quantify predictability, including 1) deriving a quantitative relation between predictability of the first and second kinds; 2) showing how differences in predictability can be decomposed in two dramatically different ways, facilitating physical interpretation; and 3) proposing a sample estimate of mutual information whose significance can be tested using standard techniques.

Abstract

Evaporative fraction (EF; the ratio of latent heat flux to the sum of the latent plus sensible heat fluxes) can be measured in the field to an accuracy of about 10%. In this modeling study, the authors try to determine to what accuracy soil moisture must be known in order to simulate surface energy fluxes within this observational uncertainty and whether there is a firm relationship between the variabilities of soil moisture and surface turbulent energy fluxes. A relationship would provide information for planning the future measurement of soil moisture, the design of field experiments, and points of focus for soil model development. The authors look for relationships in three different land surface schemes using results and ancillary integrations in the Global Soil Wetness Project.

It is found that the variation of evaporative fraction as a function of soil moisture is consistent among the models and within subsets of vegetation type. In forested areas, there is high sensitivity of EF to soil moisture variations when soils are dry and there is little sensitivity in moderate to wet soils. Where vegetation is sparser, there is a more gradual decrease of EF sensitivity with a decrease in soil moisture. Bare soil desert areas behave similarly to sparsely vegetated areas but with lower peak EF. Tundra regions have a unique behavior, probably because evaporation is limited more by a lack of radiant energy at high latitudes. The results suggest that accuracy in the measurement or model simulation of soil moisture is most critical within the drier portion of the range of variability of soil moisture. It also is more important over sparsely vegetated areas, for which evapotranspiration is dependent on moisture in a shallower column of soil.