• Moncrieff, M. W., , Waliser,, D. E., , Miller, M. J., , Shapiro, M. A., , Asrar, G. R. & , Caughey, J., 2012: Multiscale convective organization and the YOTC virtual global field campaign. Bull. Amer. Meteor. Soc., 93, 11711187.

    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., and Coauthors, 2012: The “year” of tropical convection (May 2008 to April 2010): Climate variability and weather highlights. Bull. Amer. Meteor. Soc., 93, 11891218.

    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 22 22 16
PDF Downloads 10 10 9

Progress and direction in tropical convection research: YOTC International Science Symposium

View More View Less
  • 1 National Center for Atmospheric Research, Boulder, Colorado
  • 2 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California
  • 3 THORPEX International Programme Office, WMO, Geneva, Switzerland
© Get Permissions
Full access

CORRESPONDING AUTHOR: Mitchell W. Moncrieff, NCAR, Climate and Global Dynamics Division, 1850 Table Mesa Drive, Boulder, CO 80305 E-mail: moncrief@ucar.edu

CORRESPONDING AUTHOR: Mitchell W. Moncrieff, NCAR, Climate and Global Dynamics Division, 1850 Table Mesa Drive, Boulder, CO 80305 E-mail: moncrief@ucar.edu

What: More than 100 scientists and graduate students met to discuss recent progress and new directions in understanding, simulating, and approximating the multiple scales and phenomena associated with tropical convection.

When: 16–19 May 2011

Where: Beijing, China

The Year of Tropical Convection (YOTC), jointly coordinated by the World Meteorological Organization (WMO) World Climate Research Programme (WCRP) and the World Weather Research Programme (WWRP)/THORPEX, exploits the vast pool of existing observations, high-resolution assimilation and modeling, and theoretical developments. The main objective is to advance our capabilities in weather forecasting and climate prediction with a focus on tropical convection, its multiscale organization, and interactions up to the global scale. In May 2011, the YOTC International Science Symposium, in collaboration with the Eighth AMY Workshop (see the appendix for acronyms), addressed key aspects of this challenge. Both meetings were graciously hosted by the China Meteorological Administration (CMA), Beijing, China. There were over 100 attendees, including about 20 graduate students and early career researchers supported by the WWRP, WCRP, GEO, CMA, and IAP. The symposium consisted of six sessions each introduced by three review talks, with five coordinated poster sessions and five discussion periods. Highlights of each of the session themes are summarized below. The complete program and copies of the review talks and posters are online (at www.ucar.edu/yotc).


The latent heat exchanges during phase changes of water in the atmosphere make substantial contributions to the vertical heating profile, which in turn drives atmospheric motion on various time and space scales. In the tropics, convection plays a key modulating role of these latent heat exchanges and is organized into multiscale cloud systems involving coherent circulations, a fundamental property of motion notably in a sheared atmosphere. Compared to smallscale cumulus convection, organized cloud systems have garnered little attention in parameterizations for global climate models (GCMs). In particular, mesoscale convective systems (MCS; ~100 km, ~hours to days), which are recognized to be building blocks of larger-scale organization, are absent de facto from climate models because present parameterizations do not represent organized dynamics and the model resolution is inadequate to explicitly represent (simulate) the cloud systems. In contrast, global cloud system–resolving models (GCSRMs) simulate mesoscale to large-scale cloud organization with considerable visual realism. Addressing organized tropical convection in a global context lies at the heart of the YOTC project (Moncrieff et al. 2012).

The European Centre for Medium-Range Weather Forecasts (ECMWF) provided YOTC with a database of high-resolution analysis, deterministic forecast data, and a unique set of physical diagnostics for the “year” (May 2008–April 2010). Analogous products are available from the National Aeronautics and Space Administration (NASA) and National Oceanic and Atmospheric Administration (NOAA). With the fortuitous timing of the YOTC period relative to ENSO, these databases sample La Niña and El Niño conditions and an Arctic oscillation. Moreover, in 2008 there was a near-record-setting wet North American monsoon, a severe monsoon drought in India in 2009, and a number of high-impact tropical cyclone (TC) events. Tropical wave activity included African easterly waves (AEWs), Madden–Julian oscillations (MJOs), and cases of interacting convectively coupled equatorial waves, each associated with extreme rainfall. The weather and climate highlights of the YOTC period are summarized by Waliser et al. (2012).

The absence of a “one-stop shop” for the efficient procurement and analysis of multisensor satellite data impedes research and the improvement of prediction models. The application of the NASA Goddard Giovanni satellite data framework to the YOTC project (YOTC-GS) greatly improves access via easy to use web-based tools. The YOTC-GS L2 portal facilitates access to level-2 orbital data (e.g., swath, curtain), with the YOTC-GS L3 portal providing access to level-3 orbital data (e.g., uniform grid). Satellite products available through YOTC-GS include TRMM, AIRS, QuikSCAT, AMSR-E, MODIS, etc. For detailed information, see online (at http://disc.sci.gsfc.nasa.gov/YOTC and www.ucar.edu/yotc/data.html).

Accurate representation of the diurnal cycle is crucial for weather and climate prediction. The diurnal cycle over continents is strongly modulated by propagating organized systems initiated over mountains, which are poorly represented in GCMs. Parameterization improvements have been assisted by CSRMs, such as the diurnal cycle and frequency of precipitation, ice and liquid water amounts, convection closure and triggering, and cloud variability in radiation parameterizations.


During the La Niña conditions of YOTC, the MJOs were short lived and confined to the Indian Ocean and Maritime Continent, although devastating weather occurred (e.g., February 2009 floods in northern Australia). The unusually fast eastward propagation of the MJO in April 2009 involved Kelvin waves. During the prevailing El Niño conditions, the two successive MJOs of October 2009–February 2010 propagated well into the Pacific. Convectively coupled equatorial waves occurred throughout (Waliser et al. 2012).

The dynamics of the boreal winter MJO is represented analytically as a neutrally stable interaction between planetary-scale lower-tropospheric moisture anomalies and planetary-scale convective activity. This is associated with the slow eastward phase speed, the unusual energy dispersion, and horizontal vortex structure of the MJO envelope. Theory and CSRMs show that the amplitude of the MJO is modulated by mesoscale to synoptic-scale convective organization. However, a challenge is finding clear observational support and improved prediction of these features. Distinguishing whether the organization is upscale or downscale is difficult to directly observe. In mature MJOs, important aspects include the vertical tilting associated with the first and second baroclinic vertical modes of diabatic heating, which is relatively well discerned from observations, and the third mode associated with organized convective momentum transport (CMT).

In a GCM with superparameterization, cloud-resolving models are embedded in global models in place of standard parameterizations. Superparameterization improves atmospheric variance and large-scale cloud organization, including the MJO. Coupling to the ocean typically produces further improvement. In aquaplanet simulations, stronger MJOs tend to occur with warmer sea surface temperatures, with implications for climate change. It is not understood why some aquaplanet simulations produce atmospheric equatorial superrotation while others do not. Observations, simulation, and theory all show that increased lower-tropospheric moisture enhances the strength of the MJO. Shutting down the evaporation of stratiform rain from organized convection can turn an MJO into a fast Kelvin wave. Down-gradient cumulus CMT slows the MJO and organized up-gradient transport is associated with the westerly jet in the lower troposphere.

The following questions and issues were discussed and suggest further research: (i) What causes moistening ahead of the MJO, and what is the relative role of congestus and shallow convection? (ii) Do multiscale models based on three cloud types (congestus, deep convection, and stratiform) provide an acceptable conceptual MJO, model simulation, and forecast, or is something still missing (e.g., momentum transport by organized convection)? What needs to be done observationally and computationally to evaluate recent progress in the theory and explicit simulation of the MJO? (iii) Does progress in our basic understanding of the boreal winter MJO translate to the summer MJO, or are these substantially different phenomena? (iv) The role of momentum transport by organized convection requires detailed investigation.


In the mid-1990s, previously robust correlations between seasonal precipitation in West Africa and TC activity in the Atlantic unexpectedly broke down. This implies a nonstationary correlation across time scales. In the years when the environmental conditions in the main TC development region (10°–20°N) are favorable for enhanced tropical cyclogenesis, correlations between West African monsoon rainfall and Atlantic TC activity are weaker than in years when conditions are less favorable. When conditions are less favorable, AEWs importantly affect cyclogenesis. Favorable large-scale conditions for tropical cyclogenesis and abundant West Africanrainfall have not been present since the mid-1990s. This raises the basic question: What is the role oforganized convection in triggering AEWs and in TC initiation?

Explicit TCs in high-resolution (25-km grid) climate models need to be carefully evaluated. These models usually produce too many intense storms, underestimate the decrease of storm intensity from category 5 to 4 and cluster into groups. An issue is the model's TC sensitivity to the convection parameterization. When deep convection is suppressed, storms at all levels of intensity get stronger.

Superparameterization and high-resolution global and limited domain models call for a unified approach to cloud-microphysics parameterization. Successful development and evaluation may be facilitated via multisensor satellite analysis/constraints and possibly through more judicious use of satellite simulators. The effects of resolution on TCs; multiscale interaction in cyclogenesis; and modulation of TCs by ENSO, easterly waves, and the MJO are active areas of research.

In the discussion session, considerable attention was given to resolution issues. In global models, higher horizontal resolution does not automatically generate more realistic TCs, which tend to be more similar at different resolutions than in agreement with observations. Considering that a horizontal grid of ~100 m is required to fully resolve moist convection, ~1-km grid models (i.e. CSRMs and superparameterization) do not obviate cumulus parameterization. Compared to number and tracks, progress in TC intensity has been slow. The relative roles of wind shear, moisture, and vortices and the effects of vertical resolution are incompletely understood. Environmental conditions strongly affect TCs in climate prediction while vortices are crucial forweather forecasting.


Tropical Cyclone Fengshen in 2008 was well simulated by NICAM with a 3.5-km grid. Comparison with field observations confirmed that Fengshen's genesis involved MJO, synoptic wave, and mesoscale convective systems influences. Evaluation of cloud properties using satellite data showed that the effective radius of hydrometeors in upperlevel clouds is overestimated, and ice-water content is smaller than observed. An MJO simulated by NICAM with a 7-km grid showed that the organized convective momentum transport was dominated by second and third baroclinic effects.

The high-resolution regional simulation of the April 2009 MJO, as part of the U.K. Cascade project, examines convective organization, scale interaction, and convective parameterization improvements. Higher resolution does not necessarily improve the MJO, in contrast to enhanced low-level vertical mixing. Explicit convection in a 4-km grid simulation exaggerates rainfall, underestimates the convective organization, and degrades the larger-scaleorganization but improves the diurnal cycle over the Maritime Continent. Three-dimensional turbulent mixing improves eastward propagation, moistens the lower troposphere, and retains the MJO amplitude beyond 10 days.

In the ECMWF Integrated Forecasting System, the effects of a variable adjustment time scale in the deep convection closure and an entrainment rate proportional to environmental moisture improved the MJO. Some problems remain over the Maritime Continent and with a slowing propagation in longer predictions. An ensemble of 46-day hindcasts shows the predictive skill of the MJO is now about 25 days. Hindcasts with the specified extratropics relaxed to reanalysis shows improved MJO forecast skill, indicating the importance of two-way interaction between tropical convection and the extratropics.

The discussion period highlighted a new multimodel collaborative effort, supported by the YOTC MJOTF and the GASS, on the vertical structure and diabatic processes of the MJO. The experimental framework utilizes global models with and without ocean coupling, superparameterized GCMs, GCSRMs, and nested regional models. The following will be performed: 20-year simulations to examine MJO variability in climate models; 20-day hindcasts to examine performance as a function of 1–20-day lead time; and 2-day hindcasts to examine the two successive MJO events that occurred during October 2009–February 2010.


Observational analysis linked the extreme midlatitude blocking associated with the Russian heat wave during summer 2010 to the extreme Pakistan flood in October that year. These two examples of linked weather–climate extremes, although they occurred after the “year,” garnered significant interest in regard to their dynamical origin. Global reanalysis showed a packet of Rossby wave trains linking events in the 60°–70°N zone with anomalous rainfall over Pakistan. The Bay of Bengal was the primary water vapor source.

There was some stern advice for coupled GCM predictions. The importance of running ensembles was emphasized, with the precaution that ocean-coupled GCM forecasts appear to be overly persistent and the “noise variance” is much smaller than the “forecast error variance” within the ensembles. Predictability gains are evidently more realizable in the tropics than in the extratropics. It was suggested that the MJO is primarily an atmospheric phenomenon.

An overview was given of atmospheric rivers, narrow belts of high moisture emanating from the tropics that cause extreme precipitation on the U.S. West Coast and other areas around the globe. The MJOs enhance the East Asian jet and extend it across the Pacific. When the MJO is in the Indian Ocean, Rossby wave dispersion benefits the so-called pineapple express atmospheric rivers. Our knowledge of these events and their prediction is incomplete.

The discussion period questioned the predictability of the extreme Pakistan flood in view of the difficulties models experience with extreme blocking. Regarding ensemble prediction, if an ensemble does not have adequate spread, it can be misleading. Introduction of higher resolution, superparameterization, and/or stochastic variability may have the desirable effect of increasing the ensemble spread and making the spread more comparable to forecast error. In GCMs, the large-scale circulation may not be properly linked to precipitation patterns and latent heating. By directly linking mesoscale and large-scale dynamics, organized convection may alleviate that problem and enable a key question to be addressed: Is getting the MJO right crucial for ENSO and its complex global effects?


The Asian–Australian monsoon system directly affects the world's most populous regions. It involves three distinct onset regions: Bay of Bengal, South China Sea, and South Asia. Improved prediction of the MJO is critical to skillful forecasts of the Asian–Australian monsoon. Its effect on monsoon onset and breaks is a crucial link in the seamless chain of interactions ranging from low-frequency variability such as ENSO to high-impact synoptic systems such as TCs, down to the diurnal cycle. Fundamental issues include the following: 1) a lack of accepted theories for basic features, such as characteristic frequency and spatial scales; poleward versus zonal propagation; amplification mechanisms; dependency on ocean and land surface coupling; and interannual modulation and 2) the inability of models to reliably represent many observed features, such as the intraseasonal variability (ISV) of the monsoon.

Observations show that interseasonal variability during boreal summer differs from that in winter, especially the northeastward-propagating tilted rainbands. These bands are a Rossby wave–mediated response to rain anomalies over the Maritime Continent modulated by the summertime sheared mean state. The atmospheric dynamics affecting monsoon breaks include (i) 25-day, wavenumber-three intraseasonal variability emanating from the eastern Pacific, amplified by interaction with the MJO in the western Pacific, and (ii) circulations sparked by black-carbon aerosol heating of the 1–2-km layer that feedback on active-break transitions via local circulations. More information on monsoon research and prediction is available online (at www.clivar.org/organization/aamp/aamp.php and www.ucar.edu/yotc/mjo.html), and more information on AMY and the Eighth AMY Workshop is also available online (at www.wcrp-amy.org).


TheYOTC project addresses tropical convection and its global interactions in a comprehensive manner. The focus on organized convection at the intersection of weather and climate includes the multiscale properties of the MJO where rapid progress is being made. The application of these findings to improve climate models is a priority. Notable in this regard are the synergistic uses of satellite data resources, highresolution analyses, and state-of-the art modeling frameworks, including the evaluation of climate model performance using the initialized hindcast approach. A communitywide aspect of this synergistic framework is underway as the YOTC MJO Task Force and GASS collaborative project Vertical Structure and Diabatic Processes of the MJO: A Global Multi-model Evaluation Project (see www.ucar.edu/yotc/mjo.html).


This workshop would not have occurred without the excellent hospitality and assistance of the CMA and IAP, particularly Guoxiong Wu, Xiaofeng Xu, Jianping Li, and Yihong Duan, and Jenny Lin for her organizational skills. Thanks are also due to WCRP, WWRP/THORPEX, and GEO for their generous funding support. NASA, NOAA, and the National Science Foundation (NSF) are acknowledged for their support of the YOTC Project Office through the US-THORPEX Executive Committee. DEW's contribution was carried out on behalf of the Jet Propulsion Laboratory, California Institute of Technology, under a contract with NASA. NCAR is sponsored by NSF.


  • Moncrieff, M. W., , Waliser,, D. E., , Miller, M. J., , Shapiro, M. A., , Asrar, G. R. & , Caughey, J., 2012: Multiscale convective organization and the YOTC virtual global field campaign. Bull. Amer. Meteor. Soc., 93, 11711187.

    • Search Google Scholar
    • Export Citation
  • Waliser, D. E., and Coauthors, 2012: The “year” of tropical convection (May 2008 to April 2010): Climate variability and weather highlights. Bull. Amer. Meteor. Soc., 93, 11891218.

    • Search Google Scholar
    • Export Citation


AEW: African easterly wave

AIRS: Atmospheric Infrared Sounder

AMSR-E: Advanced Microwave Scanning Radiometer for Earth Observing System (EOS)

AMY: Asian monsoon years

CMA: China Meteorological Administration

CSRM: Cloud system–resolving model

CMT: Convective momentum transport

MJOTF: Madden–Julian Oscillation Task Forcep>ENSO: El Niño–Southern Oscillation

GASS: Global Atmospheric System Study

GCM: Global climate model

GEISV: Intraseasonal (10–70 day) variabilityO: Group on Earth Observations

GCSRM: Global cloud system–resolving model

IAP: Institute of Atmospheric Physics

MCS: Mesoscale convective system

MJO: Madden–Julian oscillation

MODIS: Moderate Resolution Imaging Spectroradiometer

NICAM: Nonhydrostatic Icosahedral Atmospheric Model

QuikSCAT: Quick Scatterometer

THORPEX: The Observing System Research and Predictability Experiment

TC: Tropical cyclone

TRMM: Tropical Rainfall Measuring Mission

WCRP: World Climate Research Programme

WWRP: World Weather Research Programme

YOTC: Year of Tropical Convection