Search Results

You are looking at 1 - 10 of 26 items for

  • Author or Editor: Robert M. Chervin x
  • Refine by Access: All Content x
Clear All Modify Search
Robert M. Chervin

Abstract

A twenty-year integration of an atmospheric general circulation model with identically evolving prescribed surface boundary conditions each year is employed to provide a measure of the interannual variability obtainable from internal atmospheric dynamics alone. In particular, the variability of seasonal mean sea level pressure and 700-mb geopotential height is considered by means of a sampled climate ensemble approach. This model-generated internal dynamics variability is assumed to be identical to that resulting from the real atmosphere if it operated without anomalous boundary conditions and is considered unpredictable since the time scales involved are beyond the traditional limits of deterministic predictability. By means of objective statistical tests, sampled model variances for these fields are compared to sampled variances of observed seasonal means (which have contributions from anomalous boundary conditions as well) for all four seasons in order to ascertain if, in an infinite population sense, the range of possible climate states is reduced without interannual external variations. These tests are applied primarily for the continental United States and secondarily to the rest of the Northern Hemisphere (from field significance considerations). Indications of potential predictability from some as yet unspecified anomalous boundary conditions for the former region and possible predictability for the latter are inferred when grid point values of the sampled model variance are declared significantly less than the observed. Within this framework, it is found that no potential predictability exists in the primary area of interest for the mean sea level pressure for any season but that some potential predictability of the mean 700-mb geopotential height is obtained for limited sections of the United States for summer and winter only. Regions of possible predictability vary with both field and season, but at least one subtropical region is usually found. The ultimate predictability of such regions has to be verified by an appropriate field significance test. Results from additional numerical experiments and analyses of observed data are generally consistent with the conclusions from the original basic internal dynamics experiment, especially over the continental United States.

Full access
Robert M. Chervin

Abstract

Within the sampled climate ensemble framework for describing the climate, objective univariate statistical tests are presented which permit a straightforward assessment of the extent to which observed and GCM simulated climates agree or differ with respect to various first- and second-moment measures (i.e., ensemble averages and standard deviations) of the climate. As an example of this approach, the vertically averaged transient beat flux, (??)¯ VT is considered as a basic climate element and ensemble averages and standard deviations of this climate element are objectively compared fox. the same number of samples from a global data set assembled by A. Oort of GFDL and from sets of 2.5 and 5° resolution realizations with a GCM developed several years ago at NCAR. It is found that the degree of agreement between observed and simulated climate is highly dependent on the geographical location, the statistical moment used as the climate measure and whether zonally averaged or grid-point values constitute the measure. Also, for (??)¯ VT there is insufficient evidence to conclude that the use of a model with finer spatial resolution results in a substantially improved climate simulation. These same objective tests can and should be applied as a powerful diagnostic tool for model validation with other sampled climate ensembles either generated by other models or assembled from future observing systems.

Full access
Robert M. Chervin

Abstract

Traditionally the climate simulated by an atmospheric general circulation model (GCM) has been presented in terms of individual monthly or seasonal averages from a single realization. A more complete description of GCM simulated climate is possible by means of sampled climate ensembles consisting of sets of independent, finite time-span realizations. As an example of this sampled climate ensemble approach, estimates of first- and second-moment climate statistics are presented for grid-point data from GCM simulated climate ensembles produced by perpetual January and July versions of a 2.5° horizontal resolution GCM developed several years ago at the National Center for Atmospheric Research. Estimates of ensemble averages and standard deviations are included for time averages and for time-lagged autocorrelations for several different lags. Also included as an integral measure of persistence for these same sampled ensembles are estimates of the characteristic time (T 0) between effectively independent sample values.

Full access
Robert M. Chervin

Abstract

Increasing concern over possible anthropogenic impact on climate has led to an awareness that straightforward diagnostic procedures are necessary to measure climate and climate change in computer model experiments. Since the best documented (and most predictable) climate change is the extreme seasonal change from January to July, an obvious first application of any such set of procedures would be to determine if an atmospheric general circulation model (GCM) is capable of producing measurably different climates from a prescribed seasonal change in external forcing. Toward this end, objective statistical tests are applied to various measures of the climate to determine the extent to which sampled climate ensembles produced by January and July versions of a 5° horizontal resolution GCM developed several years ago at the National Center for Atmospheric Research differ. It is shown that while ensemble averages and standard deviations of globally averaged, time-averaged precipitation are not significantly different, significant regional differences in both first- and second-moment statistics do result from the imposed seasonal changes in surface and radiative forcing. Of course, these same objective statistical tests can and should be applied to determine the extent to which observed and simulated climate ensembles agree or differ.

Full access
Robert E. Dickinson
and
Robert M. Chervin

Abstract

A detailed analysis is presented of the sensitivity of the 12-layer tropospheric/stratosphoric NCAR global circulation model (GCM) on a zonal, global and regional basis to changes in global radiative balance due to an atmospheric burden of 10 ppb of chlorofluoromethanes (CFM's). Large changes in both the amplitude and phase of Northern Hemisphere winter planetary waves are an apparent consequence of the prescribed change. Also, large and statistically significant changes in the regional distribution of mean surface temperature and precipitation patterns result. However, many of the regional details of the model's response appear to be sensitive to the change in ocean surface temperature, which was imposed in the model run with CFM added. Hence, any more definitive indication of the climatic impact of CFM's must await the development of physically reasonable, interactive, coupled ocean-atmosphere models.

Full access
Warren M. Washington
and
Robert M. Chervin

Abstract

January and July experiments were performed with the NCAR general circulation model (GCM) to assess the potential climatic impact of the thermal energy released from a projected United States cast coast megalopolis circa 2000 A.D. The model has six layers in the vertical and a 5° latitude-longitude horizontal resolution. The ocean surface temperatures were held fixed with respect to time in both experiments at the appropriate observed climatological values for each month. To determine the statistical significance of the model response, sets of random perturbation experiments were performed for each month to obtain a measure of the model noise level (i.e., the estimated standard deviation of monthly means). Larger surface temperature changes are found in the January thermal pollution experiment. with a maximum of 12°C in the vicinity of the beat input. Smaller but still significant changes with a maximum of 3°C are found in the July experiment. Significant changes in precipitation and soil moisture also result in the prescribed change region. However, neither experiment produces any evidence of a coherent statistically significant downstream response or “teleconnection” over the Atlantic Ocean or Europe.

Although these experiments are not complete climate change experiments, in that the ocean surface temperatures and sea ice distributions are not permitted to respond to the inputed waste heat, they do demonstrate the sensitivity of a current “state of the art” GCM to such surface forcing. Furthermore, the necessity of considering different seasons in performing climatic impact studies is made apparent by the vastly different model response in the January and July experiments with the identical prescribed change in surface forcing.

Full access
Klaus M. Weickmann
and
Robert M. Chervin

Abstract

The seasonal cycle of the global wind field is documented for both a decadal set of analyses from the National Meteorological Center (NMC) and an extended term integration of a research version of the Community Climate Model developed at the National Center for Atmospheric Research. Composite eigenvector analysis is used to establish the dominant three dimensional coherent structures characteristic of the datasets while gridpoint harmonic analysis provides evidence of the extent to which these structures describe conventional seasonal modes. These quantitative indicators of spatial and temporal variance form a stringent measure of model performance with respect to seasonal variation. The model appears to be far more successful at capturing the annual harmonic contained in the NMC analyses than the semiannual harmonic. This discrepancy may be related to the absence of the requisite tropical forcings due to either inadequate parameterizations of certain physical processes or the lack of interannual variability in the model's boundary forcings, or both. Further numerical experimentation is likely to help resolve this issue.

Full access
Robert M. Chervin
and
Leonard M. Druyan

Abstract

A coarse-mesh, global climate model developed at the Goddard Institute for Space Studies (GISS) has been used to assess the influence of ocean surface temperature (OST) gradient and continentality on the Walker circulation. The basic model climate was established by a five-year integration in which the prescribed seasonal cycle in OST distribution was identical for each year. In the model climate, the Walker circulation is characterized in the zonal plane by three pairs of clockwise and counterclockwise cells to the troposphere.

Three separate winter experiments were performed in which the normal west-to-east OST gradients in the tropical Pacific were replaced by a uniform distribution in the band from 8°N to 16°S. Each experiment was characterized by OSTs set at the warmest, coldest, or mean temperatures in the band. The model response features statistically significant changes in the intensity of the various cells and branches with small shifts in the east-west extent. The overall structure in the zonal plane for the experiments with the coldest or mean temperatures, however, remained unchanged. A major disruption of the six-cell structure did result for the experiment with the warmest temperature and resultant net heat source.

The fourth prescribed changed experiment involved the replacement of the South American continent by an ocean with the OSTs linearly interpolated from the eastern Pacific to the western Atlantic. In this case, a dramatic change in the structure of the Walker circulation also took place as the upward branch over South America was reduced sufficiently to eliminate the corresponding counterclockwise cell and thereby allow two clockwise cells to merge into one large cell. The Hadley cell was less intense and shifted northward with the South American continent removed.

In summary, these experiments with the GISS model seem to indicate that both continentality and OST gradient are important as forcing mechanism of the overall structure of the Walker circulation and the intensity of the individual cells. The details of the forcing, however, are likely to be different for the two mechanisms.

Full access
David D. Houghton
and
Robert M. Chervin

Abstract

Vertically-averaged meridional transports of westerly momentum are analyzed in sampled ensembles of January simulations of an NCAR GCM and an equivalent ensemble of five years of observational January data according to a simple time-domain decomposition. Ensemble averages and standard deviations are compared in terms of both zonally-averaged and grid-point presentations for the steady and transient flux components highlighting the relative characteristics of the fundamental time-domain elements. Results from 5 and 2.5° horizontal resolution versions of the model demonstrate the impact of truncation error on model simulations of these flux statistics.

Comparing grid point measures constitutes a more stringent model performance evaluation since regional differences between observed and simulated transports often are found to he considerably larger than zonally-averaged differences. Such regional considerations also reveal substantial differences between model and observations in the location and orientation of transport maxima and minima. Typically the transient flux component is smaller in the model simulations than in the observations although there are some regional exceptions. The steady flux component, however, is generally larger in the model simulations (particularly the 2.5° version) than in the observations and is affected more than the transient component by resolution changes. Analysis of the estimated standard deviations of the flux components shows that the model's inherent variability is typically at least a factor of two lower than the observed interannual variability with substantial regional differences.

Full access
Robert M. Chervin
and
Stephen H. Schneider

Abstract

General Circulation Models (GCM's) of the earth's atmosphere have been frequently used to test different mechanisms for climate change. Typically, these tests involve comparing the statistics of the model with a prescribed change of a variable, system parameter or boundary condition, to the statistics from an unperturbed control case. However, it is not uncommon for the prescribed change experiment statistics to be comparable to statistics from a random perturbation experiment. Consequently, it is essential to determine the inherent “noise climatology” of GCM's in order to distinguish between signal and noise in climate experiments. We have examined the time-averaged response of the NCAR GCM to random perturbations in the initial conditions, while leaving all boundary conditions fixed. The dependence on time averaging interval of the noise level of a number of globally- and zonally-averaged GCM variables has been computed and gives an indication of how long to time average in order to reduce noise levels by a given amount. Additionally, we have found that it is important to delay for several simulated weeks the process of compiling climatological statistics from perturbed runs. Furthermore, the global distribution of noise levels reveals that certain regions are more prone to inherent variability for a given variable than are other areas. Also, we show that analysis of noise characteristics can be a useful diagnostic tool. However, different random perturbations do not reproduce the same noise level distribution, which implies that a Monte Carlo approach with more independent samples may he necessary for a more definite determination of the noise levels of GCM-generated statistics. Unfortunately, generating more samples means using more computer time, and that can be a fairly imposing barrier to the use of a GCM in climate experiments.

Full access