Search Results

You are looking at 1 - 9 of 9 items for

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

Abstract

Cloud-resolving model simulations of radiative–convective equilibrium are conducted in both two and three dimensions (2D and 3D) to examine the effect of dimensionality on the equilibrium statistics. Convection is forced by a fixed imposed profile of radiative cooling and surface fluxes from fixed temperature ocean.

In the control experiment, using the same number of grid points in both 2D and 3D and a zero mean wind, the temperature and moisture profiles diverge considerably after a few days of simulations. Two mechanisms are shown to be responsible for this. First, 2D geometry causes higher perturbation surface winds resulting from deep convective downdrafts, which lead to a warmer, moister boundary layer and a warmer tropospheric mean temperature state. Additionally, 2D geometry encourages spontaneous large-scale organization, in which areas far away from convection become very dry and thus inhibit further convection there, leading to a drier mean atmosphere.

Further experiments were conducted in which horizontal mean winds were applied, adopting both constant and sheared vertical profiles. With mean surface winds that are of the same magnitude as downdraft outflow velocities or greater, convection can no longer increase mean surface fluxes, and the temperature differences between 2D and 3D are greatly reduced. However, the organization of convection still exists with imposed wind profiles. Repeating the experiments on a small 2D domain produced similar equilibrium profiles to the 3D investigations, since the limited domain artificially reduces surface wind speeds, and also restricts mesoscale organization.

The main conclusions are that for modeling convection that is highly two-dimensionally organized, such as squall lines or Walker-type circulations over strong SST gradients, and for which a reasonable mean surface wind exists, it is possible that a 2D model can be used. However, for random or clustered convection, and especially in low wind environments, it is highly preferable to use a 3D cloud model.

Full access
Adrian M. Tompkins and Adeyemi A. Adebiyi

Abstract

Daily precipitation retrievals from three algorithms [the Tropical Rainfall Measuring Mission 3B42 rain product (TRMM-3B42), the Climate Prediction Center morphing technique (CMORPH), and the second version (RFEv2) of the Famine Early Warning System (FEWS)] and CloudSat retrievals of cloud liquid water, ice amount, and cloud fraction are used to document the cloud structures associated with rainfall location and intensity in the West African monsoon. The different rainfall retrieval approaches lead to contrasting cloud sensitivities between all three algorithms most apparent in the onset period of June and July. During the monsoon preonset phase, CMORPH produces a precipitation peak at around 12°N associated with upper-level cirrus clouds, while FEWS and TRMM both produce rainfall maxima collocated with the tropospheric–deep convective cloud structures at 4°–6°N. In July similar relative displacements of the rainfall maxima are observed. Conditional sampling of several hundred convection systems proves that, while upper-level cirrus is advected northward relative to the motion of the convective system cores, the reduced cover and water content of lower-tropospheric clouds in the northern zone could be due to signal attenuation as the systems there appear to be more intense, producing higher ice water contents. Thus, while CMORPH may overestimate rainfall in the northern zone due to its reliance on cloud ice, TRMM and FEWS are likely underestimating precipitation in this zone, potentially due to the use of infrared based products in TRMM and FEWS when microwave is not available. Mapping the CloudSat retrievals as a function of rain rate confirms the greater sensitivity of CMORPH to ice cloud and indicates that high-intensity rainfall events are associated with systems that are deeper and of a greater spatial scale.

Full access
F. Di Giuseppe and A. M. Tompkins

Abstract

To relate the error associated with 1D radiative calculations to the geometrical scales of cloud organization and/or in-cloud optical inhomogeneities, a new idealized methodology, based on a Fourier statistical technique, has been developed. Three-dimensional cloud fields with variability over a selected range of horizontal spatial scales and consistent vertical structure can be obtained and controlled by a small number of parameters, which relate directly to the dynamical and thermodynamical meteorology of the situation to be examined. This initial study deals with marine stratocumulus. Two experiments are conducted: an overcast situation and a broken cloud case with maximum cloud cover of 80%. For each experiment, five cloud fields are generated with the dominant organizational scale changing from 1.4 to 22 km, while all other quantities—such as cloud cover, cloud liquid water, and total water variance—remain constant. For each scene, three radiative calculations are performed for solar zenith angles of 0° and 60°: a plane parallel (PP) calculation, similar to that commonly implemented in general circulation models; an independent pixel approximation (IPA); and a full 3D calculation. The “PP bias” (IPA-PP) is used to assess cloud optical homogeneity approximation, while the “IPA bias” (3D-IPA) measures the impact of horizontal photon transport.

For the overcast scenes, the neglect of horizontal photon transport was found to be unimportant, and the IPA calculation gives accurate results. For the broken cloud case, this was only true for clouds with dominant horizontal spatial scales exceeding 10 km. With a scale of 2 km or less, the IPA bias in reflectivity, transmissivity, and absorption could exceed 5%. This indicates that even for shallow cloud systems, cloud geometry can play an important role. The sign of the bias depends critically on the solar angle, with IPA over- (under) estimating reflectivity for high (low) sun angles.

The PP bias in reflectivity was also found to be around 5% for both cases, comparable to the IPA bias, and smaller than previous estimates for this cloud type. Additional sensitivity tests prove this to be due to the vertical cloud structure. Vertically resolving the subcloud adiabatic liquid profile leads to a more opaque cloud upper boundary, reducing photon penetration into the cloud layer, and thus also PP biases. Additionally, for the broken cloud case it was found that changes in cloud fraction with height are translated by cloud overlap rules into effective horizontal variability in the liquid water path, further reducing biases. Taking a vertically uniform slab with identical integrated properties led to much larger PP biases comparable to previous estimates. Thus, models with sufficient vertical resolution are likely to suffer from smaller PP biases than previously estimated.

Full access
F. Di Giuseppe and A. M. Tompkins

Abstract

Conflicting claims have been made concerning the magnitude of the bias in solar radiative transfer calculations when horizontal photon transport is neglected for deep convective scenarios. The difficulty of obtaining a realistic set of cloud scenes for situations of complex cloud geometry, while certain characteristics such as total cloud cover are systematically controlled, has hindered the attempt to reach a consensus. Here, a simple alternative approach is adopted. An ensemble of cloud scenes generated by a cloud resolving model are modified by an idealized function that progressively alters the cirrus anvil coverage without affecting the realism of the scene produced. Comparing three-dimensional radiative calculations with the independent column approximation for all cloud scenes, it is found that the bias in scene albedo can reach as much as 22% when the sun is overhead and 46% at low sun angles. The bias is an asymmetrical function of cloud cover with a maximum attained at cirrus anvil cloud cover of approximately 30%–40%. With a cloud cover of 15%, the bias is half its maximum value, while it is limited for coverage exceeding 80%. The position of the peak occurs at the cloud cover coinciding with the maximum number of independent clouds present in the scene. Increasing the cloud cover past this point produces a decrease in the number of isolated clouds because of cloud merging, with a consequential bias reduction.

With this systematic documentation of the biases as a function of total cloud cover, it is possible to identify two contributions to the total error: the geometrical consequences of the effective cloud cover increase at low sun angles and the true 3D scattering effect of photons deviating from the original path direction. An attempt to account for the former geometrical contribution to the 1D bias is made by performing a simple correction technique, whereby the field is sheared by the tangent of the solar zenith angle. It is found that this greatly reduces the 1D biases at low sun angles. Because of the small aspect ratio of the cirrus cloud deck, the remaining bias contribution is small in magnitude and almost independent of solar zenith angle.

Full access
Frédéric Vitart, Steve Woolnough, M. A. Balmaseda, and A. M. Tompkins

Abstract

A set of five-member ensemble forecasts initialized daily for 48 days during the Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment period are performed with the ECMWF monthly forecasting system in order to assess its skill in predicting a Madden–Julian oscillation (MJO) event. Results show that the model is skillful in predicting the evolution of the MJO up to about 14 days, but the amplitude of the MJO is damped after a few days of integration. In addition, the model has some deficiencies in propagating the MJO through the Maritime Continent. The same experiment framework is used to quantify the impacts of changing the model physics, the ocean model, the atmospheric horizontal resolution, and the initial conditions on the skill of the monthly forecasting system. Results show that there is a scope for extending the skillful range of the operational monthly forecasting system to predict the evolution of the MJO by at least a week. This is achieved by using an improved cloud parameterization together with a better representation of the mixing of the upper ocean. An additional set of experiments suggests that degrading the quality of the initial conditions (by using the 15-yr ECMWF Re-Analysis instead of the 40-yr ECMWF Re-Analysis) significantly degrades the skill of the model to predict an MJO event and that increasing the horizontal resolution of the atmospheric mode had only a minor impact on the MJO forecasts. In addition, results show that there is a significant sensitivity to the initial perturbations of the ensemble members, and therefore, targeting perturbations on the MJO could improve the skill of the monthly forecasting system. While the propagation of the MJO was sensitive to most of the changes described in this paper, only the change in cloud parameterization improved the strength of the MJO. The propagation of the MJO over the Maritime Continent remains an issue.

Full access
A. N. Ross, A. M. Tompkins, and D. J. Parker

Abstract

Gravity-current models have been used for many years to describe the cold pools of low-level air that are generated by cumulonimbus precipitation. More recently, it has been realized that surface fluxes of heat and water vapor can be important in modifying these flows, through turbulent mixing of buoyancy by convection, and through direct modification of the cold pool buoyancy. In this paper, simple models describing the role of surface fluxes in depleting the negative buoyancy of a gravity current and the consequences of this for the flow dynamics are discussed.

It is pointed out that the depletion of cold pool buoyancy by surface fluxes is analogous to the depletion of buoyancy in a turbidity current through particle sedimentation, and in one regime of parameter values the analogy is exact. This analogy allows one to use simple flow models that have been tested extensively against laboratory experiments on turbidity currents. A simple “box model” and a more sophisticated shallow water model are each developed. It is shown how these models can give relatively simple expressions for cold pool “runout length” and buoyancy distributions. These runout lengths compare well with maximum cold pool sizes previously observed in cloud-resolving model simulations of unorganized tropical deep convection.

Full access

The JET2000 Project: Aircraft Observations of the African Easterly Jet and African Easterly Waves

Aircraft Observations of the African Easterly Jet and African Easterly Waves

C. D. Thorncroft, D. J. Parker, R. R. Burton, M. Diop, J. H. Ayers, H. Barjat, S. Devereau, A. Diongue, R. Dumelow, D. R. Kindred, N. M. Price, M. Saloum, C. M. Tayor, and A. M. Tompkins

Scientific background and motivation for the JET2000 aircraft observing campaign that took place in West Africa during the last week of August 2000 are presented. The Met Research Flight CI30 aircraft made two flights along the African easterly jet (AEJ) between Sal, Cape Verde, and Niamey, Niger, and two “box” flights that twice crossed the AEJ at longitudes near Niamey. Dropsondes were released at approximately 0.5°–10° intervals. The two box flights also included low-level flights that sampled north–south variations in boundary layer properties in the baroclinic zone beneath the AEJ.

Preliminary results and analysis of the JET2000 period including some of the aircraft data are presented. The JET2000 campaign occurred during a relatively dry period in the Niamey region and, perhaps consistent with this, was also associated with less coherent easterly wave activity compared to other periods in the season. Meridional cross sections of the AEJ on 28 and 29 August (after the passage of a mesoscale system) are presented and discussed. Analysis of dropsonde data on 28 August indicates contrasting convective characteristics north and south of the AEJ with dry convection more dominant to the north and moist convection more dominant to the south. The consequences of this for the AEJ and the relationship with the boundary layer observations are briefly discussed.

Preliminary NWP results indicate little sensitivity to the inclusion of the dropsonde data on the AEJ winds in European Centre for Medium-Range Weather Forecasts (ECMWF) and Met Office analyses. It is proposed that this may be due to a good surface analysis and a realistic model response to this. Both models poorly predict the AEJ in the 5-day forecast indicating the need for more process studies in the region.

Full access

Cloudnet

Continuous Evaluation of Cloud Profiles in Seven Operational Models Using Ground-Based Observations

A. J. Illingworth, R. J. Hogan, E.J. O'Connor, D. Bouniol, M. E. Brooks, J. Delanoé, D. P. Donovan, J. D. Eastment, N. Gaussiat, J. W. F. Goddard, M. Haeffelin, H. Klein Baltink, O. A. Krasnov, J. Pelon, J.-M. Piriou, A. Protat, H. W. J. Russchenberg, A. Seifert, A. M. Tompkins, G.-J. van Zadelhoff, F. Vinit, U. Willén, D. R. Wilson, and C. L. Wrench

The Cloudnet project aims to provide a systematic evaluation of clouds in forecast and climate models by comparing the model output with continuous ground-based observations of the vertical profiles of cloud properties. In the models, the properties of clouds are simplified and expressed in terms of the fraction of the model grid box, which is filled with cloud, together with the liquid and ice water content of the clouds. These models must get the clouds right if they are to correctly represent both their radiative properties and their key role in the production of precipitation, but there are few observations of the vertical profiles of the cloud properties that show whether or not they are successful. Cloud profiles derived from cloud radars, ceilometers, and dual-frequency microwave radiometers operated at three sites in France, Netherlands, and the United Kingdom for several years have been compared with the clouds in seven European models. The advantage of this continuous appraisal is that the feedback on how new versions of models are performing is provided in quasi-real time, as opposed to the much longer time scale needed for in-depth analysis of complex field studies. Here, two occasions are identified when the introduction of new versions of the ECMWF and Météo-France models leads to an immediate improvement in the representation of the clouds and also provides statistics on the performance of the seven models. The Cloudnet analysis scheme is currently being expanded to include sites outside Europe and further operational forecasting and climate models.

Full access
Adrian M. Tompkins, María Inés Ortiz De Zárate, Ramiro I. Saurral, Carolina Vera, Celeste Saulo, William J. Merryfield, Michael Sigmond, Woo-Sung Lee, Johanna Baehr, Alain Braun, Amy Butler, Michel Déqué, Francisco J. Doblas-Reyes, Margaret Gordon, Adam A. Scaife, Yukiko Imada, Masayoshi Ishii, Tomoaki Ose, Ben Kirtman, Arun Kumar, Wolfgang A. Müller, Anna Pirani, Tim Stockdale, Michel Rixen, and Tamaki Yasuda
Open access