Observation, Theory and Modeling of Atmospheric Variability

The following sections are included:

• CONTENTS

• PREFACE

• List of Reviewers

• List of NIM Alumnus Authors and Academic Years Entering NIM

• ACRONYMS

This paper reviews local instability theory and its application to storm track dynamics. It begins with a brief introduction to the concepts pertinent to storm track dynamics: convective versus absolute instability, propagating versus stationary wave-packet resonance, local growth due to energy conversions versus spatial redistributions, stochastic non-modal instability, and localization due to the basic state's deformation. A brief review and relative merits of local energetics analysis, feedback diagnostics, and diagnostics using 3-D E-P flux vectors derived from pseudo-energy conservation relations are discussed.

The physical processes that are important to local instability can be delineated from a local energetics analysis. Because the availability of background energy sources is zonally inhomogeneous, the peak amplitude location becomes an intrinsic instability property of a local mode. The storm track location is determined by an optimal balance between the local growth due to energy extracted from the basic flow and the local growth due to energy redistributions. In particular, the advection always acts to spread perturbations downstream, implying the preferred location of a storm track would be downstream of the basic jet core. The stronger the advection is, the farther downstream the storm track and the slower the growth rate of perturbations would be. This explains why the domain-averaged basic wind, rather than the domain-averaged barotropic/baroclinic wind shear, is one of the most determining factors for instability of a local mode. The energy propagation associated with the ageostrophic flow also tends to cause downstream development although it may not be as dominant as the advective processes. Another important factor for determining the storm track length is the basic deformation field in the jet exit region where meridionally elongated baroclinic eddies tend to decay barotropically.

Basic theories of the El Niño-Southern Oscillation (ENSO) phenomenon are briefly reviewed in this paper. The main focus is on the dynamics of ENSO within coupled dynamic frameworks of limited complexity. The simple coupled models capture the growth mechanism related to the positive feedback of tropical ocean-atmosphere interaction proposed by Bjerknes. Yet, different considerations of coupled processes have led to different hypotheses for the phase-transition mechanisms of ENSO-like coupled oscillations. The relevance of these theories to the understanding of the nature of the complex ENSO phenomenon will be discussed.

The land surface and the atmosphere can interact with each other through exchanges of energy, water, and momentum. With the capacity of long memory, land surface processes can contribute to long-term variability of atmospheric processes. Great efforts have been made in the past three decades to study land-atmosphere interactions and their importance to long-term variability. This paper reviews studies on monthly and seasonal variability of the land-atmosphere system. Issues to be addressed include the importance of land surface processes, time scales, persistence, coupled patterns of soil moisture and precipitation, and prediction. A perspective on future studies is given.

The theory of Ertel's potential vorticity is applied to investigating the adaptation of the atmospheric circulation to diabatic heating. Results show that such atmospheric thermal adaptation depends critically on the vertical distribution of diabatic heating. Data diagnosis shows that in summer over the Tibetan Plateau, although the column-integrated convective condensation heating overwhelms diffusive sensible heating, it is the sensible heating that determines the vertical profile of the total diabatic heating over the Tibetan Plateau and forces the prominent and unique atmospheric circulation patterns over the plateau and the surrounding area.

The strong near-surface diabatic heating over the Tibetan Plateau increases the lower layer convergence and the concavity of isentropic surfaces that intersect the plateau. Such heating, in combination with surface friction, makes the Tibetan Plateau region a strong source of negative vorticity in the atmosphere. The sensible heating-induced ascent along the lateral boundaries of the Tibetan Plateau draws in the surrounding air in the lower troposphere and pumps the air out of the region into the upper troposphere. Results obtained from numerical experiments reveal the effectiveness of such a sensible heat-driven air pump (SHAP). The SHAP not only regulates the climate in the surrounding area but also affects the circulation anomalies over the Northern Hemisphere in the form of the Rossby wave rays.

Stratosphere-troposphere exchange is dominated by the meridional circulation, with upward motion in the tropics and downward motion in the polar regions. When tropospheric air enters the stratosphere across the tropical tropopause, it is dehydrated by the cold tropopause temperatures. Thus, the tropical tropopause temperatures are essential for explaining observed distribution pattern and variations of stratospheric water vapor. Investigations about the tropical tropopause also provide information about the coupling between the stratosphere and troposphere in the tropics. In this review article, the author summarizes recent research on the tropical tropopause

The tropical tropopause was found to be influenced by the stratospheric quasi-biennial oscillation (QBO), the tropospheric El Nino southern oscillation (ENSO), and the Madden Julian oscillation (MJO). The MJO in the cold point tropopause temperatures (CPT-T) shows Kelvin wave features in the deep tropics and Rossby wave features in the subtropics. The QBO in the CPT-T is mainly zonally symmetric and is associated with downward propagating temperature anomalies accompanying the QBO meridional circulation. The influence of the ENSO on the tropical tropopause shows east-west dipole and north-south dumbbell features. A cooling trend of the tropical CPT-T was found. This cooling trend could not explain the observed positive trend of stratospheric water vapor if the tropical tropopause temperature were the only factor that determines stratospheric water vapor concentration. The three-dimensional structure of the tropical “cold trap” is also discussed.

The wintertime relationship between the Northern Hemisphere sea-ice concentration, 500-hPa height, sea level pressure and 1000-500-hPa thickness is examined. The Northern Hemispheric sea ice extent exhibits a strong sensitivity to the climatic variation of atmospheric circulation anomalies. The sea-ice extent has reduced in the Barents Sea, Greenland Sea and Labrador Sea since 1990. Particularly, the reduction of sea ice extent in the Barents Sea and Greenland Sea became evident as early as 1968. The Northern Hemispheric sea ice extent also exhibits a strong signal of decadal variability except the Greenland Sea where the downward trend is more pronounced. The sea ice variability is characterized by a dipole pattern in both the Atlantic and Pacific sectors. Its temporal variability is strongly coupled to the North Atlantic Oscillation (North Pacific Oscillation) in the Atlantic (Pacific) sector. The relationship is strongest when the atmosphere leads the sea ice by 1-2 weeks.

The piecewise potential vorticity (PV) inversion technique is applied to the ERICA IOP2 storm case. The inversion results suggest that the mid and lower level diabatic condensation-derived PV anomaly contributes a major role to this cyclone's deepening, and the contribution increases during the most rapid deepening phase. It also shows that the diabatic processes appear to augment the positive contribution of the lower boundary thermal perturbation but to reduce the relative importance of the upper level PV anomaly. Sensitivity experiments reveal that, by removing part or all of the upper level positive PV anomaly from the preconditioned atmosphere but retaining the lower level perturbation and diabatic processes, the cyclogenesis might weaken substantially. For this storm, however, the impact of the upper level PV anomaly removal seems less significant relative to diabatic processes since a dry simulation results even weaker cyclogenesis. The absence of surface thermal anomaly in the initial atmosphere could severely hamper the subsequent cyclogenesis but only if there is lack of upper level PV anomaly, implying that the thermal anomaly is an internal factor and is largely controlled by the upper level PV perturbation.

This paper introduces the concept of cloud-climate feedback along with its role in the sensitivity of the climate system. It reviews available cloud feedback diagnostic methods and representative results in three-dimensional atmospheric general circulation models. A case study is presented to analyze physical processes responsible for cloud feedbacks that control the sensitivity of the NCAR Community Climate Model. The paper further discusses weaknesses of current numerical models and areas of required research on this subject.

Atmospheric convection is an important energy source for the global circulation. It has a typical spatial scale of a few kilometers to a few tens of kilometers. Thus, in global climate models (GCMs), which have a horizontal resolution of 200 to 300 kilometers, convection must be parameterized. This study reviews some widely used convective parameterization schemes in GCMs. Emphasis will be placed on the mass flux type of schemes and associated closure conditions. Recent development on a fundamental issue related to closure will be discussed. An Example will be presented to demonstrate the role of closure using the single column version of the National Center for Atmospheric Research Community Climate Model CCM3.

Understanding cloud processes and the associated interactions with their environment is crucial for better predictions of tropical climate. The cloud-resolving model is demonstrated to be a powerful tool for process studies. In this paper, cloud modeling in the tropical deep convective regime is reviewed based on the author's research work. The review focuses on model setup, cloud-radiation interaction processes, convective-radiative processes associated with the diurnal variation of tropical oceanic convection, dominant cloud microphysical processes producing precipitation, precipitation efficiency, physical processes responsible for the phase relation between surface rain rate and convective available potential energy, and the effects of precipitation on the tropical upper ocean.

The objective of this work is twofold: (1) to understand how the turbulence interacts with the microphysics to produce the turbulent liquid water flux; (2) to develop a new framework of turbulence-microphysics coupling parameterization for stratocumulus clouds. The approach is to analyze the turbulence and microphysics data produced by a coupled large-eddy simulation and bin microphysics model. The analysis demonstrates that the condensation time scale regulates the turbulence fields, because it affects the condensation fluctuation. The liquid water flux can be parameterized in terms of the down-gradient formulation, the fluxes of conservative thermodynamic variables, and both the condensation and the dominant turbulence time scales. The new parameterization explicitly parameterizes the turbulence-microphysics coupling based on a third-order turbulence closure model, and is shown to be able to simulate some key features of the coupling process.

This paper covers topics of atmospheric data analysis and assimilation algorithms, ranging from function fitting methods to three-dimensional or four-dimensional variational approaches. The author has evaluated relevant materials in linear algebra, statistics, and optimization, and discussed major features of various methods covered. Finally, the challenges and research needed in assimilating a new type of atmospheric observation are briefly summarized.

In the past decade, ensemble forecasting has developed into a major component of numerical weather prediction. With increases in computing resources, it is becoming more realistic to produce operational ensemble forecasts for compatible members with comparable resolutions in many numerical weather prediction centers around the world. Probabilistic forecasts based on a global ensemble prediction system, especially flow-dependent forecast probability distribution, which can be readily generated from an ensemble, allow for the identification of weather systems with high and low uncertainties. The potential economic benefit achieved by using ensemble probabilistic forecasts is significant when compared to that of a deterministic forecast. Among NCEP global ensemble-based applications, the relative measure of predictability is one of the excellent prediction tools used to estimate forecast uncertainty. Probabilistic quantitative precipitation forecasts can supplement short-/medium-range forecasts. However, the ensemble forecasts, like any numerical weather prediction system, are biased. The bias of ensemble forecasts is very similar to those of a deterministic forecast and comes from the imperfections of a numerical model such as initial conditions, physical parameterizations, numerical schemes, etc. The bias of ensemble forecasts can be removed, however, by applying a statistical calibration method. With the use of such a method, calibrated ensemble forecasts and ensemble-based, calibrated probabilistic forecasts can offer the possibility of bias-free products to the meteorological community and other users.

Land exchanges momentum, energy, water, aerosols, carbon dioxide and other trace gases with its overlying atmosphere. The land surface influences climate on local, regional and global scales across a wide range of timescales. This review concentrates on the rapid (i.e., seconds to seasons) biophysical and hydrological aspects of land surface processes. This paper provides the historical development of land surface models designed for short-term weather and climate studies, ranging from the early, simple “bucket” models to recent sophisticated soil-vegetation-atmosphere transfer schemes. Major research issues are reviewed by grouping into datasets, coupling to atmospheric models, component processes, and sub-grid-scale variability and scaling. Significant problems remain to be addressed, including the difficulties in parameterizing hillslope runoff, fractional snow cover, stomatal resistance, evapotranspiration, and sub-grid-scale variability and scaling. However, further progress is expected as the results of large-scale field experiments and satellite datasets are exploited.

The benefit of coupled models has been well documented for seasonal to interannual climate predictions and global climate change assessments, whereas the advantage of mesoscale coupled models in weather and ocean predictions has only recently been exploited. Observational evidence supports the notion that mesoscale atmospheric processes over the tropical ocean and the coastal waters are coupled to the underlying ocean. Experimental forecasts of coastal ocean circulation and winter weather using coupled models confirmed the improvement in forecast skills over uncoupled models. Recent studies also showed that remarkable improvement in hurricane intensity forecasts could be achieved by using coupled hurricane-ocean models. In this paper, we provide a brief review on the subject of air-sea interactions associated with mesoscale weather systems and its application in operational weather and ocean predictions. Possible directions and opportunities for future studies are discussed.

To improve summer monsoon simulation, the regional high-resolution Eta model of the National Centers for Environmental Prediction is modified and nested in the Center for Ocean-Land-Atmosphere Studies spectral global general circulation model (GCM). A traditional one-way nesting scheme is developed and used such that the lateral boundary conditions of the Eta model are derived from GCM simulations every six hours. The Eta model domain (30°–140°E, 30°S–50°N) covers the whole Asian monsoon region, which includes the Indian, East Asian, and Southeast Asian monsoons. Both the GCM and the Eta model are integrated continuously from mid-April to the end of September to simulate the monsoon rainfall of 1987, an El Niño year, and 1988, a La Niña year, with the prescribed seasonally varying sea surface temperature. Three separate runs are made for each year with atmospheric conditions for April 14, 15, and 16. The ensemble rainfall means of the three simulations for both years are calculated for the GCM and Eta model and compared to in situ observations. Overall, the Eta model simulation shows much-improved seasonal precipitation mean and patterns, as well as intraseasonal and interannual variability as compared to its parental GCM simulations.

We investigate the effects of forcing data errors on the calibration and uncertainty estimates of the Chameleon Surface Model (CHASM) land surface model. We use Bayesian Stochastic Inversion (BSI) that is based on Bayes theorem, importance sampling, and Very Fast Simulated Annealing to search for optimal parameters and estimate the posterior probability density function (PPD) for three experiments with varying levels of forcing uncertainty. Since the complete representation of the PPD is impossible in a highly multidimensional parameter space, marginal PPDs and their maximum values are used to describe uncertainties and sensitivities in the parameters of the CHASM land surface model for five biomes representative of tropical forests, tropical pastures, midlatitude grasslands, midlatitude forests, and semiarid grasslands. The results show that forcing data errors have little effect on the calibration of the most important parameters and the estimation of their PPDs, although forcing data errors have significant effect on the selection of non-dominant optimal parameters and estimation of their PPDs. The CHASM land surface model predictions are insensitive to reasonable forcing errors for all sites except Loobos site. The Loobos example suggests that the effect of forcing errors is at least vegetation and climate dependent.

Radiative heat exchange is the major energy source in the Earth's middle atmosphere and most models of planetary atmospheres. In this article, the author will start with the basic radiative transfer equation and its formal solution from the perspective of radiative heating and cooling rate calculations in numerical models for the middle atmosphere and planetary atmospheres. The exposition of the radiative transfer theory will then be presented along three major lines that lead to the goal of heating and cooling rate calculations in a clear-sky atmosphere. First, the basic quantum concepts of the microscopic interactions between photons and particles and the physical mechanisms of absorption and emission line profiles will be summarized. Second, the traditional approaches of band models associated with frequency integration will be briefly reviewed with a special emphasis on the correlated-k distribution method. Third, the physical nature of the source function associated with non-local thermodynamic equilibrium processes will be discussed. Finally, a few selected applications of radiative transfer theory in the middle atmosphere and planetary atmospheres will be presented in the context of how the objective of high accuracy and efficiency is accomplished while solving various types of radiative transfer problems.

This article provides a tutorial for remote sensing of cloud liquid/ice water path and precipitation with satellite passive microwave measurements. First, the emission and scattering signatures in microwave measurements, and their connection with atmospheric hydrological variables are explained, followed by a description of the theoretical bases for retrieving cloud and precipitation properties with microwave observations. The retrieval methods are then briefly reviewed, and some details are provided for those algorithms developed by the author. The products derived from the microwave measurements are utilized to determine the global distribution of cloud and precipitation, assess the variability of these variables, and understand physical and radiative processes. Some challenging issues in satellite microwave remote sensing are also discussed.

This overview outlines the theoretical basis of polarimetric radiative transfer and its applications in the field of Earth science. The generic radiative transfer equation described here is applicable for the polarization and intensity in the microwave, infrared and visible portions of the spectrum. The equation is solved analytically to calculate the radiance and derivative of the radiance with respect to geophysical parameters. Both the radiance and its derivative are essential in data assimilation and in the application of remote sensing to numerical forecast models. This overview surveys the numerous applications of polarimetric signatures, such as the derivation of sea surface wind vector, sea ice parameters, long-wave radiation at the surface, and aerosol properties.

This paper reviews the current state of knowledge, advances and challenges in short-wave earth radiation budget (ERB) studies and the cloud absorption anomaly (CAA) debate. The ERB issues deal exclusively with the solar energy disposition between the atmosphere, clouds and the surface. The ERB and its disposition have been derived from surface observations, satellite remote sensing and general circulation modeling. Major sources of uncertainties and discrepancies between observations and modeling are highlighted. Reported discrepancies between the ERB data sets obtained by various means are discussed, especially within the context of the recent debate concerning the CAA. This debate is documented thoroughly and critically in four stages: before and after the mid-1990s, and following two dedicated field experiments: the Atmospheric Radiation Measurement Enhanced Shortwave Experiment (ARESE I and II). An attempt is made to shed light on the causes of some controversial findings. It is now clear that the CAA is largely an artifact which does not emerge from a carefully designed closure test using the state-of-the-art radiative transfer models.

Solar cycle effects on stratospheric ozone mixing ratio and total column ozone are studied using 23 years (1979-2001) of Solar Backscatter Ultraviolet (SBUV) data measured by Nimbus-7 and the National Oceanic and Atmospheric Administration (NOAA) satellites. It is shown that the solar cycle has the strongest signal in the upper stratospheric ozone variation, and that the ozone sensitivity to the solar cycle is comparable to that derived from the 27-day solar rotational effect. Stratospheric ozone response to the 27-day solar ultraviolet variation is also studied in three 1000-day periods with strong, moderate and weak solar ultraviolet variations respectively. The temperature effect on ozone variations is removed (or separated) by partial correlation and multiple regression methods. The correlation coefficients, ozone sensitivities and coherency squares are compared for the different periods and for cases with or without temperature effect.

This article introduces an education-orientated web page about the application of Tropical Rainfall Measuring Mission (TRMM) data in the atmospheric research. We write that on-line page with a two-fold purpose. First, it exhibits examples of the TRMM data usage in the atmospheric study. These examples provide users a means of grasping the concepts of the scientific information in the TRMM data quickly and easily. Secondly, the contents of the sample TRMM data are described and visually depicted. This is intended to familiarize users with TRMM data contents. These examples are picked up from TRMM science reports in both professional journals and distributed web sites. Rather than providing detailed descriptions of these examples, we integrate them into a brief overall page for public access. Six topics are listed at the time of this writing, including: (1) TRMM improves rainfall forecast; (2) TRMM improves rainfall data assimilation; (3) tropical cyclones: new views; (4) cloud and precipitation formation; (5) TRMM for monsoon study; and (6) TRMM for El Niño and La Niña study. Each topic covers two aspects, scientific application and sample data description. In addition, a complete list of TRMM data products, data search and order system, and data processing tools are briefly described. The web page address is at: http://eosdata.gsfc.nasa.gov/CAMPAIGN_DOCS/hydrology/TDST_SCI/sci_main.html

A brief history of precipitation from satellite passive microwave remote sensing and a theoretical basis for retrieval of precipitation from passive microwave measurements have been described. Two multi-frequency microwave-based physical inversion rain retrieval algorithms are given as examples of satellite-based rain retrieval algorithms. Some applications of rainfall from satellite measurements, especially from the Tropical Rainfall Measuring Mission (TRMM), have been summarized. Discussions on available latent heating retrieval algorithms indicate that the near future TRMM latent heating products would be reasonable. Future efforts on rain/latent heating retrievals are discussed. The quality of both precipitation and latent heating derived from satellite passive microwave measurements is expected to be better in the era of Global Precipitation Measurement mission.

Aerosol is becoming a central theme in the climate research arena, due to many new findings concerning their significant direct and indirect effects on climate (e.g. by altering temperature, cloud, radiation and precipitation) and to the large uncertainties in our estimates of aerosol forcing on climate. Despite the large loading and complex properties of East Asian aerosols, our knowledge of these aerosols and their climatic effects is so meager that they arguably present the last frontier in aerosol and climate research. While their climate effects are notably strong, the magnitude and mechanisms of their influence are far from being clear. Findings concerning how aerosols interact with energy and water cycles in other regions of cleaner environment may not be valid here. More attention needs to be focused on Asian aerosols in order to examine the existing aerosol-climate paradigms and to explore new ones. This paper provides an overview of Chinese aerosols in terms of their physical, chemical and optical properties and their potential impact on regional climate. Both anthropogenic and natural aerosols are addressed. General discussions concerning aerosol observation methods, research tools and approaches are also given. Findings by Chinese scientists are also reviewed to provide the state-of-the-art of Chinese aerosol research.

The measurements obtained from satellite microwave sensors have been increasingly utilized in numerical weather prediction models. In this study, we reviewed a comprehensive method of using a variety of microwave products for hurricane model initialization. In this method, atmospheric temperature profiles are retrieved from the Advanced Microwave Sounding Unit (AMSU) and are then utilized to derive the rotational winds for the hurricane vortex by solving the nonlinear balance equation. In addition, the divergent winds associated with latent heat release are derived from the Omega equation based on the AMSU rain rate as a diabatic heating source. Atmospheric moisture profiles are iteratively retrieved from the AMSU derived total precipitable water. Furthermore, for hurricane simulations over oceans, the sea surface temperature (SST) is an important component in the overall hurricane initialization. This parameter can be uniquely derived from lower frequencies of satellite microwave sensors. The operational microwave SST products have been made available from the Tropical Rainfall Measuring Mission microwave imager.

The retrieved temperatures in four tropical cyclones shown reasonable warm-core structures, and corresponded well with the storm intensities. Trajectory analyses were performed using model output data from a 5-day explicit simulation of Hurricane Bonnie. It was found that the upper-level westerly flow converged into the west eyewall, and then produced middle-level descending warming in the west to the south eyewall. The frontogenesis function was frontolytic in the west eyewall, with the contributions from the deformation term at low levels and from the tilting term at middle and upper levels. It was shown that upper-level strong westerly inflow was a key factor leading to the development of the partial eyewall.

In this paper, the censoring technique is used to deal with block-structured Markov renewal processes. Two probabilistic measures, the R- and G-measures, are defined and a censoring invariant property for the R- and G-measures is obtained. The RG-factorization for the transition probability matrix is derived based on the Wiener-Hopf type equations constructed in this paper. For a Markov renewal process of GI/G/1 type, the transition probability matrix consists of two sequences of matrices, one is referred to as the boundary sequence and the other the repeating sequence. The RG-facterization of the double transformations for the repeating matrix sequence and four matrix inequalities of the double transformations for the boundary matrix sequence are given.

In this paper, the authors will investigate the maximum likelihood estimator of fractional autoregressive integrated moving-average time series parameter. Recently, Dahlhaus considered the likelihood function of the processes. Substituting the unknown mean µ by any consistent estimator, Dahlhaus has shown that the maximum likelihood estimator of the time series parameter ? is consistent and asymptotically normal. The authors will estimate µ and ? simultaneously by maximizing the full likelihood function and prove the consistency and asymptotic normality for the estimator of ?. The results also hold for the long range dependence processes.

Scrambled quasi-Monte Carlo quadrature proposed by Owen (1995) is a hybrid of Monte Carlo and quasi-Monte Carlo methods, which combines the best of these two methods for integration. This article reports some of main results on the scrambled quadrature.

Through examples in channel flow and solid combustion, this paper studies the asymptotic behavior of small disturbances as they evolve in time in the presence of linear instability. Weakly unstable problems where the parameter marginally crosses the neutral value are analyzed and multiple scale expansions are used to derive the dominant evolution equation that governs the weakly nonlinear behavior of solution for long times. For the shallow water waves in an inclined open channel, the evolution equation turns out to be an integro-partial differential equation of first order that can be solved numerically in conjunction with the jump condition that follows from the exact bore conditions. For the channel flow which includes internal dissipation, long wave approximation is made and the evolution equation is a generalized Kuramoto-Sivashinsky equation. For the solid combustion problem, the onset of linear instabilities is studied, and the evolution equations that govern the amplitudes of the normal mode solution are of the Landau-Stuart type. In the first example, the asymptotic theory predicts the solution accurately for both the transient and quasi-steady phases. For the second and third examples, formation of coherent structures is analyzed.

The recently developed empirical mode decomposition (EMD) method is applied to developing an algorithm for the continuous wavelet transform based on the fast Fourier transform that greatly reduces edge effects for a finite data series. By introducing an anti-symmetric padding on both sides of the intrinsic mode functions derived from the EMD method, the spectral characteristics of the original data series near the edges are appropriately preserved in the local power spectrum. Numerical experiments show that the proposed algorithm based on the anti-symmetric padding is better than the traditional zero-padding technique in deriving a more accurate power spectrum of various scales near the edges.

The following sections are included:

• AUTHOR INDEX