Richard B. Rood

Stratosphere-troposphere exchange

In the past, studies of stratosphere-troposphere exchange of mass and chemical species have mainly emphasized the synoptic- and small-scale mechanisms of exchange. This review, however, includes also the global-scale aspects of exchange, such as the transport across an isentropic surface (potential temperature about 380 K) that in the tropics lies just above the tropopause, near the 100-hPa pressure level. Such a surface divides the stratosphere into an “overworld” and an extratropical “lowermost stratosphere” that for transport purposes need to be sharply distinguished. This approach places stratosphere-troposphere exchange in the framework of the general circulation and helps to clarify the roles of the different mechanisms involved and the interplay between large and small scales. The role of waves and eddies in the extratropical overworld is emphasized. There, wave-induced forces drive a kind of global-scale extratropical “fluid-dynamical suction pump,” which withdraws air upward and poleward from the tropical lower stratosphere and pushes it poleward and downward into the extratropical troposphere. The resulting global-scale circulation drives the stratosphere away from radiative equilibrium conditions. Wave-induced forces may be considered to exert a nonlocal control, mainly downward in the extratropics but reaching laterally into the tropics, over the transport of mass across lower stratospheric isentropic surfaces. This mass transport is for many purposes a useful measure of global-scale stratosphere-troposphere exchange, especially on seasonal or longer timescales. Because the strongest wave-induced forces occur in the northern hemisphere winter season, the exchange rate is also a maximum at that season. The global exchange rate is not determined by details of near-tropopause phenomena such as penetrative cumulus convection or small-scale mixing associated with upper level fronts and cyclones. These smaller-scale processes must be considered, however, in order to understand the finer details of exchange. Moist convection appears to play an important role in the tropics in accounting for the extreme dehydration of air entering the stratosphere. Stratospheric air finds its way back into the troposphere through a vast variety of irreversible eddy exchange phenomena, including tropopause folding and the formation of so-called tropical upper tropospheric troughs and consequent irreversible exchange. General circulation models are able to simulate the mean global-scale mass exchange and its seasonal cycle but are not able to properly resolve the tropical dehydration process. Two-dimensional (height-latitude) models commonly used for assessment of human impact on the ozone layer include representation of stratosphere-troposphere exchange that is adequate to allow reasonable simulation of photochemical processes occurring in the overworld. However, for assessing changes in the lowermost stratosphere, the strong longitudinal asymmetries in stratosphere-troposphere exchange render current two-dimensional models inadequate. Either current transport parameterizations must be improved, or else, more likely, such changes can be adequately assessed only by three-dimensional models.

Multidimensional Flux-Form Semi-Lagrangian Transport Schemes

Abstract An algorithm for extending one-dimensional, forward-in-time, upstream-biased, flux-form transport schemes (e.g., the van Leer scheme and the piecewise parabolic method) to multidimensions is proposed. A method is also proposed to extend the resulting Eulerian multidimensional flux-form scheme to arbitrarily long time steps. Because of similarities to the semi-Lagrangian approach of extending time steps, the scheme is called flux-form semi-Lagrangian (FFSL). The FFSL scheme can be easily and efficiently implemented on the sphere. Idealized tests as well as realistic three-dimensional global transport simulations using winds from data assimilation systems are demonstrated. Stability is analyzed with a von Neuman approach as well as empirically on the 2D Cartesian plane. The resulting algorithm is conservative and upstream biased. In addition, it contains monotonicity constraints and conserves tracer correlations, therefore representing the physical characteristics of constituent transport.

An assimilated dataset for Earth science applications

The Data Assimilation Office at NASAs Goddard Space Flight Center is currently producing a multiyear gridded global atmospheric dataset for use in climate research, including tropospheric chemistry applications. The data, which are being made available to the scientific community, are well suited for climate research since they are produced by a fixed assimilation system designed to minimize the spinup in the hydrological cycle. By using a nonvarying system, the variability due to algorithm change is eliminated and geophysical variability can be more confidently isolated. The analysis incorporates rawinsonde reports, satellite retrievals of geopotential thickness, cloud-motion winds, and aircraft, ship, and rocketsonde reports. At the lower boundary, the assimilating atmospheric general circulation model is constrained by the observed sea surface temperature and soil moisture derived from observed surface air temperature and precipitation fields. The available output data include all prognostic variables...

Three-dimensional radon 222 calculations using assimilated meteorological data and a convective mixing algorithm

The distribution of 222Rn is simulated using a three-dimensional chemistry and transport model driven by assimilated data. The multiyear calculation is the first to use meteorological data from the Goddard Earth Observing System data assimilation system (GEOS-1 DAS). In addition, this calculation is the first to use moist convection and boundary layer parameters directly from a DAS to calculate mixing via moist convection and turbulence. Previously, these quantities have been derived after the assimilation procedure. Model output and data are compared at sites selected to evaluate model performance in a range of dynamic environments. Simulated afternoon boundary layer concentrations are within 30% of observed concentrations. Simulated nighttime concentrations are ≈50% of measured values although the bias is greatly reduced (and even disappears at Socorro, New Mexico) when nocturnal sub-grid-scale turbulent mixing is suppressed. Continental profiles are “C-like”; a result that is consistent with the moist convective algorithms tendency to move material from the planetary boundary layer (PBL) to the upper troposphere directly. CTM-calculated 222Rn at Bermuda closely matches observations even during periods of frontal passage showing that the atmospheric circulation is accurate and that ventilation of the continental PBL is realistic. Model-calculated 222Rn concentrations in the marine upper troposphere are consistent with observations in the eastern Pacific and near Darwin.

Transport-induced interannual variability of carbon monoxide determined using a chemistry and transport model

Transport-induced interannual variability of carbon monoxide (CO) is studied during 1989–1993 using the Goddard chemistry and transport model (GCTM) driven by assimilated data. Seasonal changes in the latitudinal distribution of CO near the surface and at 500 hPa are captured by the model. The annual cycle of CO is reasonably well simulated at sites of widely varying character. Day to day fluctuations in CO due to synoptic waves are reproduced accurately at remote North Atlantic locations. By fixing the location and magnitude of chemical sources and sinks, the importance of transport-induced variability is investigated at CO-monitoring sites. Transport-induced variability can explain 1991–1993 decreases in CO at Mace Head, Ireland, and St. Davids Head, Bermuda, as well as 1991–1993 increases in CO at Key Biscayne, Florida. Transport-induced variability does not explain decreases in CO at southern hemisphere locations. The model calculation explains 80–90% of interannual variability in seasonal CO residuals at Mace Head, St. Davids Head, and Key Biscayne and at least 50% of variability in detrended seasonal residuals at Ascension Island and Guam. Upper tropospheric interannual variability during October is less than 8% in the GCTM. Exceptions occur off the western coast of South America, where mixing ratios are sensitive to the strength of an upper tropospheric high, and just north of Madagascar, where concentrations are influenced by the strength of offshore flow from Africa.

Upper-Tropospheric Water Vapor from UARS MLS

Abstract Initial results of upper-tropospheric water vapor obtained from the Microwave Limb Sounder (MLS) on the Upper Atmosphere Research Satellite (UARS) are presented. MLS is less affected by clouds than infrared or visible techniques, and the UARS orbit provides daily humidity monitoring for approximately two-thirds of the earth. Best results are currently obtained when water vapor abundances are approximately 100–300 ppmv, corresponding to approximately 12-km height in the Tropics and 7 km at high latitudes. The observed latitude variation of water vapor at 215 hPa is in good agreement with the U.K. Universitiess Global Atmospheric Modelling Project model. The ability to observe synoptic-scale features associated with tropopause height variations is clearly illustrated by comparison with the National Aeronautics and Space Administration Goddard Space Flight Center assimilation model. Humidity detrainment streams extending from tropical convective regions are also observed.

A three‐dimensional simulation of the ozone annual cycle using winds from a data assimilation system

The wind fields from the NASA Goddard stratospheric data assimilation procedure are used in a three-dimensional chemistry and transport model to produce an ozone simulation for the year September 11, 1991 to September 10, 1992. Photochemical production and loss are taken from the Goddard two-dimensional model. The calculated ozone is compared with observations from the total ozone mapping spectrometer (TOMS) onboard Nimbus 7 and the microwave limb sounder on the upper atmospheric research satellite. Although the model total ozone is about 50 Dobson units (DU; =2.69 × 10−16 molecules cm−2) lower than TOMS in the tropics and up to 70 DU higher than TOMS in middle to high latitudes, the simulated ozone fields reproduce many of the features in the observations. Even at the end of this integration, the synoptic features in the modeled total ozone are very similar to TOMS observations, indicating that the model maintains realistic values for the horizontal and vertical gradients, at least in the lower stratosphere. From this good comparison between model and observations on timescales ranging from days to months, we infer that the transport driven by the assimilated wind fields closely approximates the actual atmospheric transport. Therefore the assimilated winds are useful for applications which may be sensitive to the lower stratospheric transport.

Two‐dimensional and three‐dimensional model simulations, measurements, and interpretation of the influence of the October 1989 solar proton events on the middle atmosphere

The very large solar proton events (SPEs) which occurred from October 19 to 27, 1989, earned substantial middle-atmospheric HOx and NOx constituent increases. Although no measurements of HOx increases were made during these SPEs, increases in NO were observed by rocket instruments which are in good agreement with calculated NO increases from our proton energy degradation code. Both the HOx and the NOx increases can cause ozone decreases; however, the HOx-induced ozone changes are relatively short-lived because HOx species have lifetimes of only hours in the middle atmosphere. Our two-dimensional model, when used to simulate effects of the longer-lived NOx, predicted lower-stratospheric polar ozone decreases of greater than 2% persisting for one and a half years past these SPEs. Previous three-dimensional model simulations of these SPEs (Jackman et al., 1993) indicated the importance of properly representing the polar vortices and warming events when accounting for the ozone decreases observed by the solar backscattered ultraviolet 2 instrument two months past these atmospheric perturbations. In an expansion of that study, we found that it was necessary to simulate the November 1, 1989, to April 2, 1990, time period and the November 1, 1986, to April 2, 1987, time period with our three-dimensional model in order to more directly compare to the stratospheric aerosol and gas experiment (SAGE) II observations of lower stratospheric NO2 and ozone changes between the end of March 1987 and 1990 at 70°N. Both the NOx increases from the October 1989 SPEs and the larger downward transport in the 1989–1990 northern winter compared to the 1986–1987 northern winter contributed to the large enhancements in NO2 in the lower stratosphere observed in the SAGE II measurements at the end of March 1990. Our three-dimensional model simulations predict smaller ozone decreases than those observed by SAGE II in the lower stratosphere near the end of March 1990, indicating that other factors, such as heterogeneous chemistry, might also be influencing the constituents of this region.

The GEOS ozone data assimilation system: Specification of error statistics

A global three-dimensional ozone data assimilation system has been developed at the Data Assimilation Office of the NASA Goddard Space Flight Center. The Total Ozone Mapping Spectrometer (TOMS) total ozone data and the Solar Backscatter Ultraviolet/2 (SBUV/2) partial ozone profile observations are assimilated. The assimilation, into an off-line ozone transport model, is done using the global Physical-space Statistical Analysis Scheme. This system became operational in December 1999. A detailed description of the statistical analysis scheme and, in particular, of the forecast- and observation-error covariance models is given. A new global anisotropic horizontal forecast-error correlation model accounts for a varying distribution of observations with latitude. Correlations are largest in the zonal direction in the tropics where data are sparse. Forecast-error variance is assumed to be proportional to the ozone field. The forecast-error covariance parameters were determined by maximum-likelihood estimation. The error covariance models are validated using χ2 statistics. The analysed ozone fields in the winter 1992 are validated against independent observations from ozone sondes and the Halogen Occultation Experiment (HALOE). The difference between the mean HALOE observations and the analysis fields is less than 10% at pressure levels between 70 and 0.2 hPa. The global root-mean-square difference between TOMS observed and forecast values is less than 4%. The global root-mean-square difference between SBUV observed and analysed ozone between 50 and 3 hPa is less than 15%.

Application of a monotonic upstream-biased transport scheme to three-dimensional constituent transport calculations

Abstract The application of van Leers scheme, a monotonic, upstream-biased differencing scheme, to three-dimensional constituent transport calculations is shown. The major disadvantage of the scheme is shown to be a self-limiting diffusion. A major advantage of the scheme is shown to be its ability to maintain constituent correlations. The scheme is adapted for a spherical coordinate system with a hybrid sigma-pressure coordinate in the vertical. Special consideration is given to cross-polar flow. The vertical wind calculation is shown to be extremely sensitive to the method of calculating the divergence. This sensitivity implies that a vertical wind formulation consistent with the transport scheme is essential for accurate transport calculations. The computational savings of the time-splitting method used to solve this equation are shown. Finally, the capabilities of this scheme are illustrated by an ozone transport and chemistry model simulation.