Browse
Abstract
Tropical convection that occurs on large-enough space and time scales may evolve in response to large-scale balanced circulations. In this scenario, large-scale midtropospheric vorticity anomalies modify the atmospheric stability by virtue of thermal wind gradient balance. The convective vertical mass flux and the moisture profile adjust to changes in atmospheric stability that affect moisture and entropy transport. We hypothesize that the convection observed during the 2011 DYNAMO field campaign evolves in response to balanced dynamics. Strong relationships between midtropospheric vorticity and atmospheric stability confirm the relationship between the dynamic and the thermodynamic environments, while robust relationships between the atmospheric stability, the vertical mass flux, and the saturation fraction provide evidence of moisture adjustment. These results are important because the part of convection that occurs as a response to balanced dynamics is potentially predictable. Furthermore, the diagnostics used in this work provide a simple framework for model evaluation, and suggest that one way to improve simulations of large-scale organized deep tropical convection in global models is to adequately capture the relationship between the dynamic and thermodynamic environments in convective parameterizations.
Abstract
Tropical convection that occurs on large-enough space and time scales may evolve in response to large-scale balanced circulations. In this scenario, large-scale midtropospheric vorticity anomalies modify the atmospheric stability by virtue of thermal wind gradient balance. The convective vertical mass flux and the moisture profile adjust to changes in atmospheric stability that affect moisture and entropy transport. We hypothesize that the convection observed during the 2011 DYNAMO field campaign evolves in response to balanced dynamics. Strong relationships between midtropospheric vorticity and atmospheric stability confirm the relationship between the dynamic and the thermodynamic environments, while robust relationships between the atmospheric stability, the vertical mass flux, and the saturation fraction provide evidence of moisture adjustment. These results are important because the part of convection that occurs as a response to balanced dynamics is potentially predictable. Furthermore, the diagnostics used in this work provide a simple framework for model evaluation, and suggest that one way to improve simulations of large-scale organized deep tropical convection in global models is to adequately capture the relationship between the dynamic and thermodynamic environments in convective parameterizations.
Abstract
An unfiltered zonal Hovmöller depiction of rainfall in the Maritime Continent (MC) reveals remarkable spatiotemporal continuity of zonally propagating disturbances with a diurnal period, which endure over multiple days and propagate faster than the individual convective storms they coupled with. This phenomenon and its sensitivity to the Madden–Julian oscillation (MJO) during the 2011/12 Dynamics of the MJO (DYNAMO) field campaign is examined here through a well-validated, convection-permitting model simulation conducted on a large domain. We find that these disturbances are zonally propagating diurnal gravity waves excited by vigorous nocturnal mesoscale convective systems over Sumatra and Borneo. These gravity waves are diurnally phase locked: their wavelength very closely matches the distance between these two islands (~1500 km), while their particular zonal phase speed (~±17 m s−1) allows them to propagate this distance in one diurnal cycle. We therefore hypothesize that these waves are amplified by resonant interaction due to diurnal phase locking. While these zonal gravity waves decouple from convection once beyond the MC, their divergent flow signature endures well across the Indian Ocean, provoking the notion that they may influence rainfall at far remote locations. The exact controls over this zonal phase speed remain uncertain; we note, however, that it is roughly consistent with diurnal offshore-propagating modes documented previously. Further study is required to tie this down, and more generally, to understand the sensitivity of these modes to background flow strength and the geography of the MC.
Abstract
An unfiltered zonal Hovmöller depiction of rainfall in the Maritime Continent (MC) reveals remarkable spatiotemporal continuity of zonally propagating disturbances with a diurnal period, which endure over multiple days and propagate faster than the individual convective storms they coupled with. This phenomenon and its sensitivity to the Madden–Julian oscillation (MJO) during the 2011/12 Dynamics of the MJO (DYNAMO) field campaign is examined here through a well-validated, convection-permitting model simulation conducted on a large domain. We find that these disturbances are zonally propagating diurnal gravity waves excited by vigorous nocturnal mesoscale convective systems over Sumatra and Borneo. These gravity waves are diurnally phase locked: their wavelength very closely matches the distance between these two islands (~1500 km), while their particular zonal phase speed (~±17 m s−1) allows them to propagate this distance in one diurnal cycle. We therefore hypothesize that these waves are amplified by resonant interaction due to diurnal phase locking. While these zonal gravity waves decouple from convection once beyond the MC, their divergent flow signature endures well across the Indian Ocean, provoking the notion that they may influence rainfall at far remote locations. The exact controls over this zonal phase speed remain uncertain; we note, however, that it is roughly consistent with diurnal offshore-propagating modes documented previously. Further study is required to tie this down, and more generally, to understand the sensitivity of these modes to background flow strength and the geography of the MC.
Abstract
Understanding convective initiation of the Madden–Julian oscillation (MJO) remains an unmet challenge. MJO initiation has been perceived as a process starting from a convectively suppressed large-scale condition with gradual growth of shallow convection to congestus and to deep convective and stratiform systems that cover a large-scale area. During the DYNAMO field campaign over the Indian Ocean, MJO initiation was observed to start from an existing intertropical convergence zone (ITCZ) south of the equator. This raises a question of what possible role the ITCZ may play in convective initiation of the MJO. This study addresses this question through analysis of satellite observations of precipitation and a global reanalysis product. By setting several criteria, MJO and ITCZ events were objectively identified and grouped according to whether MJO initiation was immediately preceded by an ITCZ. The results demonstrate that an ITCZ is neither a necessary nor sufficient condition for convective initiation of the MJO. Nonetheless, evolution of the large-scale circulation, moisture, and convective characteristics during MJO initiation can be different with and without a preexisting ITCZ. Convective growth begins gradually before and during MJO initiation when there is a preexisting ITCZ whereas it is abrupt and slightly delayed without a preexisting ITCZ. Such differences are presumably related to the existing large-scale moist condition of the ITCZ. The results from this study suggest that there are multiple mechanisms for convective initiation of the MJO, which should be considered in theoretical understanding of the MJO.
Abstract
Understanding convective initiation of the Madden–Julian oscillation (MJO) remains an unmet challenge. MJO initiation has been perceived as a process starting from a convectively suppressed large-scale condition with gradual growth of shallow convection to congestus and to deep convective and stratiform systems that cover a large-scale area. During the DYNAMO field campaign over the Indian Ocean, MJO initiation was observed to start from an existing intertropical convergence zone (ITCZ) south of the equator. This raises a question of what possible role the ITCZ may play in convective initiation of the MJO. This study addresses this question through analysis of satellite observations of precipitation and a global reanalysis product. By setting several criteria, MJO and ITCZ events were objectively identified and grouped according to whether MJO initiation was immediately preceded by an ITCZ. The results demonstrate that an ITCZ is neither a necessary nor sufficient condition for convective initiation of the MJO. Nonetheless, evolution of the large-scale circulation, moisture, and convective characteristics during MJO initiation can be different with and without a preexisting ITCZ. Convective growth begins gradually before and during MJO initiation when there is a preexisting ITCZ whereas it is abrupt and slightly delayed without a preexisting ITCZ. Such differences are presumably related to the existing large-scale moist condition of the ITCZ. The results from this study suggest that there are multiple mechanisms for convective initiation of the MJO, which should be considered in theoretical understanding of the MJO.
Abstract
The Madden–Julian oscillation (MJO) dominates the intraseasonal variability of cloud populations of the tropical Indian and Pacific Oceans. Suppressed MJO periods consist primarily of shallow and isolated deep convection. During the transition to an active MJO, the shallow and isolated deep clouds grow upscale into the overnight hours. During active MJO periods, mesoscale convective systems occur mostly during 2–4-day bursts of rainfall activity with a statistically significant early morning peak. Yet when these rain events are separated into individual active periods, some periods do not follow the mean pattern, with the November events in particular exhibiting an afternoon peak. The radar-observed microphysical processes producing the precipitation during the major rain events of active MJO periods evolve in connection with synoptic-scale wave passages with varying influences of diurnal forcing. MJO studies that do not account for the intermittency of rainfall during active MJO phases through averaging over multiple events can lead to the misimpression that the primary rain-producing clouds of the MJO are modulated solely by the diurnal cycle.
Abstract
The Madden–Julian oscillation (MJO) dominates the intraseasonal variability of cloud populations of the tropical Indian and Pacific Oceans. Suppressed MJO periods consist primarily of shallow and isolated deep convection. During the transition to an active MJO, the shallow and isolated deep clouds grow upscale into the overnight hours. During active MJO periods, mesoscale convective systems occur mostly during 2–4-day bursts of rainfall activity with a statistically significant early morning peak. Yet when these rain events are separated into individual active periods, some periods do not follow the mean pattern, with the November events in particular exhibiting an afternoon peak. The radar-observed microphysical processes producing the precipitation during the major rain events of active MJO periods evolve in connection with synoptic-scale wave passages with varying influences of diurnal forcing. MJO studies that do not account for the intermittency of rainfall during active MJO phases through averaging over multiple events can lead to the misimpression that the primary rain-producing clouds of the MJO are modulated solely by the diurnal cycle.
Abstract
We seek to use ARM MJO Investigation Experiment (AMIE)-DYNAMO field campaign observations to significantly constrain height-resolved estimates of the parameterization-relevant, causal sensitivity of convective heating Q to water vapor q. In field data, Q profiles are detected via Doppler radar wind divergence D while balloon soundings give q. Univariate regressions of D on q summarize the information from a 10-layer time–pressure series from Gan Island (0°, 90°E) as a 10 × 10 matrix. Despite the right shape and units, this is not the desired causal quantity because observations reflect confounding effects of additional q-correlated casual mechanisms. We seek to use this matrix to adjudicate among candidate estimates of the desired causal quantity: Kuang’s matrix 
Abstract
We seek to use ARM MJO Investigation Experiment (AMIE)-DYNAMO field campaign observations to significantly constrain height-resolved estimates of the parameterization-relevant, causal sensitivity of convective heating Q to water vapor q. In field data, Q profiles are detected via Doppler radar wind divergence D while balloon soundings give q. Univariate regressions of D on q summarize the information from a 10-layer time–pressure series from Gan Island (0°, 90°E) as a 10 × 10 matrix. Despite the right shape and units, this is not the desired causal quantity because observations reflect confounding effects of additional q-correlated casual mechanisms. We seek to use this matrix to adjudicate among candidate estimates of the desired causal quantity: Kuang’s matrix
Abstract
This study examines the westward-propagating convective disturbances with quasi-2-day intervals of occurrence identified over Gan Island in the central Indian Ocean from mid- to late October 2011 during the Dynamics of the Madden–Julian Oscillation (DYNAMO) field campaign. Atmospheric sounding, satellite, and radar data are used to develop a composite of seven such disturbances. Composites and spectral analyses reveal that 1) the quasi-2-day convective events comprise westward-propagating diurnal convective disturbances with phase speeds of 10–12 m s−1 whose amplitudes are modulated on a quasi-2-day time scale on a zonal scale of ~1000 km near the longitudes of Gan; 2) the cloud life cycle of quasi-2-day convective disturbances shows a distinct pattern of tropical cloud population evolution—from shallow to deep to stratiform convection; 3) the time scales of mesoscale convective system development and boundary layer modulation play essential roles in determining the periodicity of the quasi-2-day convective events; and 4) in some of the quasi-2-day events there is evidence of counterpropagating (westward and eastward) cloud systems along the lines proposed by Yamada et al. Based on these findings, an interpretation is proposed for the mechanisms for the quasi-2-day disturbances observed during DYNAMO that combines concepts from prior studies of this phenomenon over the western Pacific and Indian Oceans.
Abstract
This study examines the westward-propagating convective disturbances with quasi-2-day intervals of occurrence identified over Gan Island in the central Indian Ocean from mid- to late October 2011 during the Dynamics of the Madden–Julian Oscillation (DYNAMO) field campaign. Atmospheric sounding, satellite, and radar data are used to develop a composite of seven such disturbances. Composites and spectral analyses reveal that 1) the quasi-2-day convective events comprise westward-propagating diurnal convective disturbances with phase speeds of 10–12 m s−1 whose amplitudes are modulated on a quasi-2-day time scale on a zonal scale of ~1000 km near the longitudes of Gan; 2) the cloud life cycle of quasi-2-day convective disturbances shows a distinct pattern of tropical cloud population evolution—from shallow to deep to stratiform convection; 3) the time scales of mesoscale convective system development and boundary layer modulation play essential roles in determining the periodicity of the quasi-2-day convective events; and 4) in some of the quasi-2-day events there is evidence of counterpropagating (westward and eastward) cloud systems along the lines proposed by Yamada et al. Based on these findings, an interpretation is proposed for the mechanisms for the quasi-2-day disturbances observed during DYNAMO that combines concepts from prior studies of this phenomenon over the western Pacific and Indian Oceans.
Abstract
Reforecasts produced by the ECMWF Integrated Forecast System (IFS) were used to study heating and moistening processes associated with three MJO events over the equatorial Indian Ocean during the Dynamics of the Madden–Julian Oscillation (DYNAMO) field campaign. Variables produced by and derived from the IFS reforecast (IFS-RF) agree reasonably well with observations over the DYNAMO sounding arrays, and they vary smoothly from the western to eastern equatorial Indian Ocean. This lends confidence toward using IFS-RF as a surrogate of observations over the equatorial Indian Ocean outside the DYNAMO arrays. The apparent heat source Q 1 and apparent moisture sink Q 2 produced by IFS are primarily generated by parameterized cumulus convection, followed by microphysics and radiation. The vertical growth of positive Q 1 and Q 2 associated with the progression of MJO convection can be gradual, stepwise, or rapid depending on the event and its location over the broader equatorial Indian Ocean. The time for convective heating and drying to progress from shallow (800 hPa) to deep (400 hPa) can be <1 to 6 days. This growth time of heating and drying is usually short for convective processes alone but becomes longer when additional microphysical processes, such as evaporative moistening below convective and stratiform clouds, are in play. Three ratios are calculated to measure the possible role of radiative feedback in the MJO events: amplitudes of radiative versus convective heating rates, changes in radiative versus convective heating rates, and diabatic (with and without the radiative component) versus adiabatic heating rates. None of them unambiguously distinguishes the MJO from non-MJO convective events.
Abstract
Reforecasts produced by the ECMWF Integrated Forecast System (IFS) were used to study heating and moistening processes associated with three MJO events over the equatorial Indian Ocean during the Dynamics of the Madden–Julian Oscillation (DYNAMO) field campaign. Variables produced by and derived from the IFS reforecast (IFS-RF) agree reasonably well with observations over the DYNAMO sounding arrays, and they vary smoothly from the western to eastern equatorial Indian Ocean. This lends confidence toward using IFS-RF as a surrogate of observations over the equatorial Indian Ocean outside the DYNAMO arrays. The apparent heat source Q 1 and apparent moisture sink Q 2 produced by IFS are primarily generated by parameterized cumulus convection, followed by microphysics and radiation. The vertical growth of positive Q 1 and Q 2 associated with the progression of MJO convection can be gradual, stepwise, or rapid depending on the event and its location over the broader equatorial Indian Ocean. The time for convective heating and drying to progress from shallow (800 hPa) to deep (400 hPa) can be <1 to 6 days. This growth time of heating and drying is usually short for convective processes alone but becomes longer when additional microphysical processes, such as evaporative moistening below convective and stratiform clouds, are in play. Three ratios are calculated to measure the possible role of radiative feedback in the MJO events: amplitudes of radiative versus convective heating rates, changes in radiative versus convective heating rates, and diabatic (with and without the radiative component) versus adiabatic heating rates. None of them unambiguously distinguishes the MJO from non-MJO convective events.
Abstract
The atmospheric circulation depends on poorly understood interactions between the tropical atmospheric boundary layer (BL) and convection. The surface moist static energy (MSE) source (130 W m−2, of which 120 W m−2 is evaporation) to the tropical marine BL is balanced by upward MSE flux at the BL top that is the source for deep convection. Important for modeling tropical convection and circulation is whether MSE enters the free troposphere by dry turbulent processes originating within the boundary layer or by motions generated by moist deep convection in the free troposphere. Here, highly resolved observations of the BL quantify the MSE fluxes in approximate agreement with recent cloud-resolving models, but the fluxes depend on convective conditions. In convectively suppressed (weak precipitation) conditions, entrainment and downdraft fluxes export equal shares (60 W m−2) of MSE from the BL. Downdraft fluxes are found to increase 50%, and entrainment to decrease, under strongly convective conditions. Variable entrainment and downdraft MSE fluxes between the BL and convective clouds must both be considered for modeling the climate.
Abstract
The atmospheric circulation depends on poorly understood interactions between the tropical atmospheric boundary layer (BL) and convection. The surface moist static energy (MSE) source (130 W m−2, of which 120 W m−2 is evaporation) to the tropical marine BL is balanced by upward MSE flux at the BL top that is the source for deep convection. Important for modeling tropical convection and circulation is whether MSE enters the free troposphere by dry turbulent processes originating within the boundary layer or by motions generated by moist deep convection in the free troposphere. Here, highly resolved observations of the BL quantify the MSE fluxes in approximate agreement with recent cloud-resolving models, but the fluxes depend on convective conditions. In convectively suppressed (weak precipitation) conditions, entrainment and downdraft fluxes export equal shares (60 W m−2) of MSE from the BL. Downdraft fluxes are found to increase 50%, and entrainment to decrease, under strongly convective conditions. Variable entrainment and downdraft MSE fluxes between the BL and convective clouds must both be considered for modeling the climate.
Abstract
This study uses high-resolution rainfall estimates from the S-Polka radar during the DYNAMO field campaign to examine variability of the diurnal cycle of rainfall associated with MJO convection over the Indian Ocean. Two types of diurnal rainfall peaks were found: 1) a late afternoon rainfall peak associated with the diurnal peak in sea surface temperatures (SSTs) and surface fluxes and 2) an early to late morning rainfall peak associated with increased low-tropospheric moisture. Both peaks appear during the MJO suppressed phase, which tends to have stronger SST warming in the afternoon, while the morning peak is dominant during the MJO enhanced phase. The morning peak occurs on average at 0000–0300 LST during the MJO suppressed phase, while it is delayed until 0400–0800 LST during the MJO enhanced phase. This delay partly results from an increased upscale growth of deep convection to broader stratiform rain regions during the MJO enhanced phase. During the MJO suppressed phase, rainfall is dominated by deep and isolated convective cells that are short-lived and peak in association with either the afternoon SST warming or nocturnal moisture increase. This study demonstrates that knowledge of the evolution of cloud and rain types is critical to explaining the diurnal cycle of rainfall and its variability. Some insights into the role of the complex interactions between radiation, moisture, and clouds in driving the diurnal cycle of rainfall are also discussed.
Abstract
This study uses high-resolution rainfall estimates from the S-Polka radar during the DYNAMO field campaign to examine variability of the diurnal cycle of rainfall associated with MJO convection over the Indian Ocean. Two types of diurnal rainfall peaks were found: 1) a late afternoon rainfall peak associated with the diurnal peak in sea surface temperatures (SSTs) and surface fluxes and 2) an early to late morning rainfall peak associated with increased low-tropospheric moisture. Both peaks appear during the MJO suppressed phase, which tends to have stronger SST warming in the afternoon, while the morning peak is dominant during the MJO enhanced phase. The morning peak occurs on average at 0000–0300 LST during the MJO suppressed phase, while it is delayed until 0400–0800 LST during the MJO enhanced phase. This delay partly results from an increased upscale growth of deep convection to broader stratiform rain regions during the MJO enhanced phase. During the MJO suppressed phase, rainfall is dominated by deep and isolated convective cells that are short-lived and peak in association with either the afternoon SST warming or nocturnal moisture increase. This study demonstrates that knowledge of the evolution of cloud and rain types is critical to explaining the diurnal cycle of rainfall and its variability. Some insights into the role of the complex interactions between radiation, moisture, and clouds in driving the diurnal cycle of rainfall are also discussed.
Abstract
The November 2011 Madden–Julian oscillation (MJO) event during the Dynamics of the MJO (DYNAMO) field campaign is simulated with the Regional Atmospheric Modeling System (RAMS) cloud-resolving model to examine the relationship between precipitation and surface latent heat flux (LHFLX) for deep convective clusters within the MJO and to discern the importance of surface LHFLX for organizing MJO convection. First, a simulation similar in size to the DYNAMO northern sounding array was run with interactive surface fluxes. Composites for precipitation, surface LHFLX, wind speed, wind vectors, and near-surface specific humidity are described for various-sized convective clusters during different MJO regimes. The precipitation–LHFLX relationship generally evolves as follows for an individual cluster. About 2 h before cluster identification, the maximum LHFLX occurs upwind of maximum precipitation. As cluster identification time is approached, LHFLX and precipitation maxima become coincident. At and after the cluster is identified, maximum LHFLXs move downwind of the precipitation maximum with a local minimum in LHFLXs behind the precipitation maximum.
Sensitivity simulations with spatially homogenized LHFLXs were then run to determine the impacts of local LHFLX feedbacks on convective organization. Using area-averaged convective versus stratiform precipitation fraction and a simple convective aggregation index to quantify organization, no systematic difference in convective organization was detected between the control and sensitivity simulations, suggesting that local LHFLX variability is not important to convective organization in this model. Implications of these results are discussed.
Abstract
The November 2011 Madden–Julian oscillation (MJO) event during the Dynamics of the MJO (DYNAMO) field campaign is simulated with the Regional Atmospheric Modeling System (RAMS) cloud-resolving model to examine the relationship between precipitation and surface latent heat flux (LHFLX) for deep convective clusters within the MJO and to discern the importance of surface LHFLX for organizing MJO convection. First, a simulation similar in size to the DYNAMO northern sounding array was run with interactive surface fluxes. Composites for precipitation, surface LHFLX, wind speed, wind vectors, and near-surface specific humidity are described for various-sized convective clusters during different MJO regimes. The precipitation–LHFLX relationship generally evolves as follows for an individual cluster. About 2 h before cluster identification, the maximum LHFLX occurs upwind of maximum precipitation. As cluster identification time is approached, LHFLX and precipitation maxima become coincident. At and after the cluster is identified, maximum LHFLXs move downwind of the precipitation maximum with a local minimum in LHFLXs behind the precipitation maximum.
Sensitivity simulations with spatially homogenized LHFLXs were then run to determine the impacts of local LHFLX feedbacks on convective organization. Using area-averaged convective versus stratiform precipitation fraction and a simple convective aggregation index to quantify organization, no systematic difference in convective organization was detected between the control and sensitivity simulations, suggesting that local LHFLX variability is not important to convective organization in this model. Implications of these results are discussed.
