Abstract

The “hot tower” hypothesis requires the existence of deep cumulonimbus clouds in the deep Tropics as essential agents, which accomplish the mass and energy transport essential for the maintenance of the general circulation. As the role of the deep convective clouds has been generally accepted, the popularity of referring to these deep “hot” towers as undilute towers also has gained acceptance. This paper examines the consequences of assuming that the deep convective clouds over tropical oceans consist of undilute ascent from the subcloud layer.

Using simple applications of parcel theory, it is concluded that observed properties of typical cumulonimbus updrafts in low- to midtroposphere over tropical oceans are inconsistent with the presence of undilute updrafts. Such undilute updrafts are far more consistent with observations in severe storms of midlatitudes. The observations over tropical oceans can be hypothetically explained by assuming large dilution of updrafts by entrainment below about 500 hPa, followed by freezing of condensate. This freezing and subsequent ascent along an ice adiabat reinvigorates the updrafts and permits them to reach the tropical tropopause with the necessary high values of moist static energy, as the hot tower hypothesis requires. The large difference observed between ocean and land clouds can be explained by assuming slightly smaller entrainment rates for clouds over land. These small entrainment differences have a very large effect on updrafts in the middle and upper troposphere and can presumably account for the large differences in convective vigor, ice scattering, and lightning flash rates that are observed. It follows that convective available potential energy (CAPE) is not a particularly good predictor of the behavior of deep convection.

Using the Tropical Rainfall Measuring Mission (TRMM) to map a proxy for the most intense storms on earth between 36°S and 36°N, they are found mostly outside the deep Tropics, with the notable exception of tropical Africa.

Abstract

The thunderstorm frequency over the oceans during the Global Atmospheric Research Program Atlantic Tropical Experiment is quantified by examination of over 20 000 surface hourly observations from research ships. The overall thunderstorm frequency is one thunderstorm day per ship per month. There were many examples of intense mesoscale systems, such as squall lines, passing over the ships, extending to 13–17 km in altitude, but that nevertheless produce few reports of lightning. This reinforces the idea, based on data from other tropical ocean regions and from global satellite data, that in spite of the ubiquitous “hot towers” over tropical oceans, marine cumulonimbus product little lightning.

Climatological data from the monsoon regions of the Tropics are analyzed to reveal that during periods of onshore flow and heavy rainfall the oceanic regime of high rainfall but little lightning moves onshore. A rain-thunderstorm ratio is defined and used to characterize convective rainfall regimes as continental (relatively little) or maritime (relatively great) rainfall compared to the number of thunderstorm days. In regions such as West Africa and south Asia, the seasonal rainfall peak is actually accompanied by a thunderstorm minimum.

It is further suggested that the data support the idea, not original here, that vertical velocities in oceanic cumulonimbus clouds tend to be low compared with continental clouds. Radar data from the companion paper in this issue are also consistent with this idea. It is hypothesized that most oceanic storms have updrafts weaker than a possible threshold value, below which the supercooled liquid water, large ice particles, and ice-ice collisions are not present in the mixed-phase region in sufficient concentrations for electrification leading to lightning.

Abstract

The Line Islands Experiment, conducted on and near Palmyra, Fanning and Christmas Islands during February-April 1967, produced extensive data on disturbances of the equatorial trough zone. One disturbance which passed through the heart of the data network is analyzed in detail. This disturbance intensified rapidly just east of Fanning Island during the night of 31 March–1 April, but satellite observations show that it dissipated rapidly during the daylight hours of 1 April. The convergence-divergence patterns associated with the growth and decay of the disturbance are most intense in the lowest 500 m. Data from serial rawinsonde releases on the islands, combined with research aircraft data, are presented which demonstrate that highly unsaturated downdrafts are produced, first on the convective scale and the mesoscale, and finally becoming organized over the entire 600-km extent of the system. Cumulus development is effectively suppressed in the downdraft air, only being restored after 6–12 hr by the greatly enhanced energy flux from sea to atmosphere, and through the boundary layer. In order to produce the observed downdrafts, it is shown that the three-dimensional circulation patterns and thermodynamic processes within regions of intense convection are closely analogous to those in typical mid-latitude squall lines.

The Line Islands Experiment has resulted in unique and comprehensive data for studies of the meteorology of the equatorial Pacific. It is one of several recent field programs in tropical meteorology designed to attack the central problem of scale interactions, especially the role of convective and mesoscale systems. Some of the recent evidence is reviewed that indicates the importance of these interactions in understanding the non-steady state aspects of tropical disturbances. A variety of results from the Line Islands Experiment are summarized, with emphasis on their relevance to the planning of GARP tropical experiments.

Abstract

This is Part II of a two-part paper describing the vertical profile of radar reflectivity in GATE convective cells. Time-height radar life histories for 42 cells over three GATE days are examined, using data from the Quadra radar with 5-minute resolution. Mean profiles and plots of cell characteristics are generated, and confirm that the mean profiles in Part I are representative of the active portion of the cell lifetime. There are marked differences between the cell life histories of isolated cells and the longer-lived cells associated with mesoscale systems. In contrast to cells sampled in organized systems, the isolated cells are often of very limited vertical extent and must be dominated by the warm rain process. When forcing features exist such as gust fronts and intersecting lines of convection, they appear to dominate the generation of new convection, and isolated strong echoes are not observed.

Composite life histories for typical GATE cells are constructed. The typical radar echo forms first at an altitude of 2.5 km and reaches the surface about 5 minutes later, strongly suggesting early domination by the warm rain process. At the same time the echo top rises and the mid-to-late stages of cell lifetime involve both warm rain and ice processes. The reflectivity profiles of the longer-lived echoes change relatively little in the middle 50% of the life cycle.

Abstract

Based on the University of Utah Tropical Rainfall Measuring Mission (TRMM) precipitation feature (PF) database, tropical cyclone PFs (TCPFs) are identified for over 600 storms that reached tropical storm intensity level or above around the globe during eight TC seasons from the period of 1998–2006. Each TC season includes 6 months yr⁻¹. Six basins are considered: Atlantic (ATL), east-central Pacific (EPA), northwest Pacific (NWP), north Indian Ocean (NIO), south Indian Ocean (SIO), and South Pacific (SPA). TRMM 2A25- (precipitation radar) and 3B42- (multisatellite) derived rainfall amounts are used to assess the impact of tropical cyclone (TC) rainfall in altering the regional, seasonal, and interannual distribution of the global total rainfall during the TC seasons in the six basins. The global, seasonal, and interannual variations of the monthly rainfall inside TCPFs are presented. The fractional rainfall contributions by TCPFs are compared in different basins. The TRMM 2A25 and 3B42 retrievals are compared in terms of the rainfall contribution by TCs. After constraining TC rainfall for being within 500 km from the TC center, 2A25 and 3B42 show similar results: 1) TCs contribute, respectively, 8%–9%, 7%, 11%, 5%, 7%–8%, and 3%–4% of the seasonal rainfall to the entire domain of the ATL, EPA, NWP, NIO, SIO, and SPA basins; 2) both algorithms show that, regionally, the maximum percentage of TC rainfall contribution is located in EPA basin near the Mexico Baja California coast (about 55%), SIO close to the Australia coast (about 55%), and NWP near Taiwan (about 35%–40%); 3) the maximum monthly percentage of TC rainfall contribution is in September for the ATL basin, August and September for EPA, August for NWP, May for NIO, March for SIO, and January and February for SPA; 4) the percentage of rainfall contributed by TCs is higher during El Niño years than La Niña years for EPA and NWP basins. The trend is the reverse for ATL and NIO, and nearly neutral for SIO and SPA. However, this study does not include enough years of data to expect the findings to be representative of long-term statistics of the interannual variations.

Abstract

A comprehensive passive microwave satellite dataset is analyzed to quantify and compare the time evolution of convective properties of the pregenesis stage of developing disturbances (12 cases) and nondeveloping disturbances (3 cases), to determine whether the properties within the day prior to formation are unique, and to determine whether there is a credible connection between convection and the organization of the incipient circulation. Cases examined were the focus of recent (since 2005) field programs, and include those investigated during the triagency field programs in the Atlantic during 2010 [NASA’s Genesis and Rapid Intensification Processes (GRIP) project, the National Science Foundation (NSF)/NCAR Pre-Depression Investigation of Cloud-Systems in the Tropics (PREDICT) program, and NOAA’s Intensity Forecasting Experiment (IFEX)]. Among the properties examined (raining area, intensity, areal coverage of “strong” and “intense” convection, frequency, and proximity to the disturbance center), the results indicate that the area and frequency of rainfall within 3° are distinguishably greater in developing disturbances. Except for the fact it occurs in a more organized disturbance, there does not appear to be anything special about strong [polarization corrected temperature (PCT) ≤ 210 K] or intense (PCT ≤ 160 K) convection occurring in the day before genesis. Strong and intense convection events are observed throughout the pregenesis stage, do not necessarily increase (in intensity and area) as genesis nears, and are not necessarily very close (within 1°) to the center within a day of genesis. Likewise, while the areal coverage of strong and intense convection during the pregenesis stage is typically greater in developing disturbances, the overall intensity of convection in nondeveloping disturbances is comparable to the developing cases examined.

Abstract

High-resolution passive microwave observations within the stratiform regions of two different tropical oceanic mesoscale convective systems are investigated in detail. The observations were obtained from the Advanced Microwave Precipitation Radiometer (AMPR) during the Tropical Ocean and Global Atmosphere Coupled Ocean-Atmosphere Response Experiment. The AMPR records data at 10.7, 19.35, 37.1, and 85.5 GHz.

Data collected during one ER-2 leg over the stratiform regions of each system are examined in detail. (Each leg is well coordinated with the low-flying WP-3s.) The passive microwave observations of the two stratiform regions, with similar surface rain rates, are found to differ significantly from one another. The average 85.5-GHz brightness temperature was 30 K colder on 22 February compared with 20 February. This greater ice scattering on 22 February is found to be consistent with the simultaneous radar reflectivity profile, which shows that upper-level reflectivities are greater through a deep layer. This and comparison of the two meteorological situations suggest a more recent injection of ice particles from the adjacent active convection region on 22 February, and demonstrate that the 85.5-GHz temperatures are sensitive to these differences.

The 19.35- and 37.1-GHZ brightness temperatures were somewhat higher for a given 10.7-GHz temperature an 20 February. That system has a stronger bright band, and higher radar reflectivities below the melting level, suggesting greater vertically integrated rain water content. The radar reflectivities on 20 February decreased markedly with distance below the melting level, implying that a greater proportion of the rain evaporated. It is suggested that the brightness temperatures of the three lower frequencies were somewhat sensitive to these differences in the vertical distribution of precipitation within the rain layer between the two cases.

Indications are that many of the differences between these two stratiform regions on these two days result from differences in their stage of evolution. At the time the detailed analyses were made for the 20 February case, the stratiform region had more time to mature compared with the newly formed stratiform region sampled on 22 February.

Abstract

A tropical depression which intensified quite close to Dakar on 15 July 1974 was one of the strongest weather systems investigated during the GATE field program. A coordinated data set, consisting of satellite, radar, dropwindsonde and ship data, is analyzed to define both synoptic-scale and smaller scale event for a period of about 6 h. The wind fields and precipitation fields are presented in some detail for the time with the best data coverage, and the changes that took place for several hours on either side of that time are discussed. Some mesoscale events of considerable strength occurred. Mesoscale organization of deep convection, accompanied by strong mesoscale convergence at low levels, preceded mesoscale cyclogenesis. There was a marked mesoscale cyclonic center bordered in part by deep convective clouds that resembled the eye structure of a tropical cyclone in some respects. The large-scale thermodynamic stratification is noted to have been an important control on the distribution of deep convection within the tropical depression and on the eventual disappearance of deep convection.

Abstract

This study provides quantitative estimates of the thermodynamic and kinematic structures of the troposphere during various convective regimes observed during the Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment. The data source is the upper air soundings from six stations in the intensive flux array. A correction algorithm has been applied to the humidity data to remove biases between the stations. The data are analyzed using the nonhierarchical clustering method known as k means. Eleven thermodynamic clusters and 20 kinematic clusters are selected.

The thermodynamic clusters are grouped into four general categories based on their midtropospheric equivalent potential temperature. Deep convective activity varies with the thermodynamic structure of the environment. When the “dry intrusion” group is observed, convection is suppressed. The “fair weather” category corresponds to undisturbed periods with light winds and small mesoscale convective systems (MCSs). The largest MCSs and the majority of the rainfall occur with the “active” and “convective recovery” categories.

The kinematic clusters are also divided into four general categories based on the strength, direction, and depth of the low-level zonal flow. The timing of the clusters is related to the intraseasonal oscillation (ISO). Dry phases of the ISO are characterized by the “low-level easterly” category. During transition periods between the easterly and westerly phases of the ISO, the “calm” category is often seen. The “moderate shear westerly” group is seen just before the strongest westerlies. The majority of the clusters fall into the “strong shear westerly” group, associated with the peak westerly phase of the ISO.