1. Introduction
Vigorous deep cumulus convection in the tropics plays a pivotal role in driving the global atmospheric circulation and large-scale disturbances such as intraseasonal variability, synoptic waves, and tropical cyclones. Therefore, numerous studies have been dedicated to deepening our understanding of interaction between the convection and large-scale environmental fields and to developing conceptual models of an ensemble of convective cells, some of which have been utilized in numerical models as cumulus parameterization schemes. One of the key environmental factors that determine the intensity and behavior of convection is lower-tropospheric buoyancy (e.g., Ahmed and Neelin 2018), in which thermodynamic conditions in the atmospheric boundary layer (BL) play essential roles.
In spatial scales large enough to contain an ensemble of convective cells and time scales longer than their typical lifetime, local tendency terms in budget equations of conservative thermodynamic quantities in the BL are negligibly small compared with other terms representing physical processes that increase or decrease the quantities, and thus the tendency terms can be neglected. This is known as the BL quasi-equilibrium (BLQE; Raymond 1995) approximation. As for the BL moist static energy (MSE), there are four processes that play major roles in its budget over warm tropical ocean (Raymond 1994, 1995; Thayer-Calder and Randall 2015; de Szoeke 2018). Turbulent heat flux across the sea surface practically always imports MSE from the ocean to the BL in the tropics, as the sea surface temperature (SST) is usually high enough. At the top of the BL, we consider two processes that bring free-tropospheric air masses to the BL. Since these air masses have lower MSE than the BL air, these processes decrease the BL MSE per unit mass, or the BL MSE export. One of them is entrainment of air just above the BL top caused by turbulence in the BL. The other is convective downdrafts, which bring midtropospheric air to the BL. In addition, radiative processes usually cool the BL air and thus export MSE, although they have often been neglected in previous studies (e.g., de Szoeke 2018), as their contribution is usually smaller than that of the other processes.
Previous studies that developed conceptual models of deep cumulus convection had different views from each other as to which process plays the dominant role in the BL MSE export, as briefly reviewed by Thayer-Calder and Randall (2015). For example, the cumulus parameterization scheme developed by Arakawa and Schubert (1974) neglected the convective downdraft, while Cheng and Arakawa (1997) revised it to take this process into account. In contrast, Emanuel (1989) proposed a theory for tropical cyclone dynamics with the entrainment neglected. Raymond (1994) also neglected the entrainment in his model of tropical convection. Furthermore, Raymond (1995) argued that the contribution of the entrainment was negligibly small.
Observational results are expected to be helpful in deepening the above discussion. However, whereas numerous studies have developed the estimation method of the surface turbulent flux over the ocean (e.g., Fairall et al. 2003), only few studies have attempted to evaluate the contributions of the entrainment and convective downdraft using observational data. Recently, de Szoeke (2018) estimated the BL MSE export due to these two processes from shipborne upper-air and surface meteorological observations over the equatorial Indian Ocean in October–December 2011 as part of the Cooperative Indian Ocean Experiment on Intraseasonal Variability in the Year 2011 (CINDY2011)/DYNAMO field campaign (Yoneyama et al. 2013). The estimated MSE export due to the convective downdraft was larger than that due to the entrainment on average. Furthermore, the contribution of convective downdraft was enhanced in convectively active phases of the Madden–Julian oscillation (MJO; Madden and Julian 1971, 1972) compared with that in preceding phases.
Although de Szoeke (2018) provided valuable information for deepening our understanding of the BL MSE budget over the tropical ocean with vigorous cumulus convection, this is only a case study over a particular area in a particular period. Therefore, it seems fruitful to conduct a similar analysis applied to shipborne observations of other field campaign projects over tropical oceanic areas to accumulate quantitative information on the contributions of the two processes. Furthermore, we need to test whether the estimated downward mass fluxes of the two processes are consistent with the intensity of cumulus convection, which was not addressed in de Szoeke (2018).
One of the field campaigns considered suitable for such research is the Pacific Area Long-term Atmospheric Observation for Understanding of Climate Change in 2008 (PALAU2008; e.g., Katsumata et al. 2013) conducted over the tropical western Pacific in June 2008 by the Japan Agency for Marine–Earth Science and Technology (JAMSTEC). A major part of the campaign was shipborne observation using Research Vessel (R/V) Mirai equipped with a weather radar, whose observational data are expected to make possible the comparison of the estimated mass fluxes with convective activity around the vessel. Furthermore, a convectively active phase of the boreal summer intraseasonal oscillation (BSISO; a recent comprehensive review was provided by Kikuchi 2021), the boreal-summer counterpart of MJO, passed over the observation area during the observation period.
The purpose of this study is to analyze observational data from the PALAU2008 field campaign to estimate the contributions of entrainment and convective downdraft to the BL MSE budget. We will also discuss whether the estimated mass fluxes of the two processes are physically consistent with the convective activity around the vessel, and then examine variability associated with BSISO. The results of these analyses will be compared with those by de Szoeke (2018). The remainder of this paper is organized as follows. In section 2, the observational data analyzed and characteristics of large-scale fields during the PALAU2008 campaign are briefly explained. The estimation method used in this study is introduced in section 3. Results and discussions are presented in section 4. Finally, a summary and conclusions of this paper are given in section 5.
2. Data
The observational data analyzed in this study were obtained by the PALAU2008 field campaign project over the tropical western Pacific. As the main component of the campaign, R/V Mirai was deployed at 12°N, 135°E from 6 to 27 June 2008 to perform intensive observations (cruise ID: MR08–02). Please refer to JAMSTEC (2008) for details of the cruise. Among a variety of atmospheric and oceanographic observation items conducted on the vessel, this study analyzes surface meteorological parameters, upper-air sounding data, weather radar data, and ceilometer data. The surface meteorological parameters analyzed are horizontal wind velocity at 25 m above mean sea level (MSL), temperature and humidity at 21 m MSL, downwelling shortwave and longwave radiation and air pressure at 13 m MSL, and an intake seawater temperature at a 5-m depth, all of which have a temporal resolution of 10 min. Upper-air sounding was performed every 3 h using Vaisala RS92 sensors. Temperature, humidity, wind speed and direction, and pressure data were analyzed.
Volume scan observation using the shipborne C-band Doppler radar was performed every 10 min throughout the observation period to monitor the convective activity around the vessel. We calculated the fractional area covered by radar echo stronger than 15 dBZ at 2 km MSL within a 30-km range from the vessel, which is hereafter referred to as radar echo coverage (REC). The REC is a measure of the intensity of cumulus convection accompanied by precipitation. We also use cloud-base height data just above the vessel observed by the ceilometer every 1 min, from which we calculated the fractional time when the cloud base is detected below 1 km MSL, which is referred to as low-cloud coverage (LCC). The REC and LCC are evaluated as daily means.
In addition to the shipborne observation, 6-hourly upper-air soundings were performed at Koror (7.5°N, 134.5°E) and Yap (9.5°N, 138.1°E) weather stations. These two stations and the vessel constituted a triangle, and the upper-air data at the three sites enable us to examine budget analyses of conserved quantities such as mass in the triangle, as pioneered by Yanai et al. (1973). We will use 6-hourly area-averaged horizontal convergence estimated from these data by Katsumata et al. (2013).
Large-scale circulation fields during the campaign were characterized by northward migration of a convectively active phase of BSISO, as described in detail by Geng et al. (2011, 2013, 2014) and Katsumata et al. (2013). Figure 1a shows a time–latitude cross section of precipitation and column-integrated water vapor (CWV) averaged over 125°–145°E (including the observation area) in June 2008. In the first half of June, an area with high precipitation amount and high CWV was located south of 10°N, which then migrated northward in mid-June and arrived at the observation area around 20 June. A time–latitude cross section of zonal wind at the 850-hPa level (Fig. 1b) shows that easterly winds prevailed in the first half of June, whereas an area with westerly wind was found to the south of the northward-migrating high-precipitation area. The westerly winds lasted several days over the observation area. Then they were replaced by easterly winds again in the last few days of the observation period. On the basis of these large-scale characteristics, we divide the observation period into two equal-length subperiods (6–16 and 17–27 June) to compare the MSE budget processes between these periods. The former period will be referred to as the preactive period, as the convectively active phase of BSISO arrived just after this period. The latter period will be called the active period. Note that Katsumata et al. (2013) took a similar approach, although they omitted 16 and 17 June for some reasons.
Time–latitude cross sections of (a) precipitation (color) and CWV (contours) and (b) zonal wind at the 850-hPa level averaged over 125°–145°E. Thin and thick contours in (a) indicate 48 and 52 mm, respectively. The horizontal dashed line represents the period and latitude of the shipborne observation. The data plotted are the Global Satellite Mapping of Precipitation (GSMaP; Kubota et al. 2020) for precipitation, and the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015) for CWV and wind.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
The daily time series of convective activity around the vessel is shown in Fig. 2. There are 4 days (14, 20, 21, and 26 June) when the convection is relatively active with REC higher than 10%. The maximum in REC was recorded on 21 June, which was arguably associated with the arrival of the convectively active phase of BSISO. Averages in REC over the preactive and active periods are 2.4% and 8.4%, respectively. On the four days with high REC, LCC is also relatively high compared with that on the other days (except for 10 June). The correlation coefficient between REC and LCC is +0.57, which is statistically significant at the 99% confidence level. Interestingly, LCC is lowest in 22–24 June, just after the arrival of the convectively active phase. These time series will be compared with the entrainment and convective downdraft mass fluxes in section 4b.
Daily time series of (a) REC and (b) LCC.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
3. Method
This section describes the method for estimating the entrainment and convective downdraft mass fluxes, which is similar to that in de Szoeke (2018) but with several modifications. The estimated mass fluxes will be used to evaluate the contributions of the two processes to the BL MSE budget.
We make several simplifications to the BL characteristics, as in previous studies (e.g., Raymond 1995; de Szoeke 2018). First, we disregard variability in the depth of the BL (zb) and set it to 500 m. This value is considered typical over tropical oceanic areas (Garstang and Betts 1974). Furthermore, Geng et al. (2013) estimated zb from the upper-air sounding data during PALAU2008 and reported that its average over 6–24 June was 482 m, which is close to the value we set. Geng et al. (2013) also found a decreasing trend in zb over the course of the 19-day period at a rate of approximately 10 m day−1, which corresponds to a downward mass flux at the BL top of 0.14 g s−1 m−2. As will be shown in section 4b, this value is much smaller than the entrainment and convective downdraft mass fluxes, suggesting that disregarding the zb tendency is a good approximation. Second, we assume that there are no clouds in the BL. Actually, the ceilometer data revealed that the cloud base was rarely observed below 500 m during the observation period. We also assume that dry static energy (DSE) and water vapor mixing ratio (hereafter called mixing ratio for brevity) were vertically well mixed over the depth of the BL. Last, we assume BLQE and no horizontal advection of DSE and mixing ratio in the BL.
We will use Eqs. (1) and (2) to calculate me and mcd with other quantities in the equations estimated from observations as follows. The surface fluxes (H and E) are calculated from the surface meteorological observations using the COARE3.0 bulk flux algorithm (Fairall et al. 2003). The radiative heating R is estimated from surface meteorological observations, as summarized in the appendix. Note that, although the radiative process is often neglected in previous studies, including de Szoeke (2018), we will demonstrate that considering it leads to a more physically consistent estimation of the mass fluxes.
For estimating the DSE and mixing ratio of the BL, entrainment, and convective downdraft air masses, we basically follow de Szoeke et al. (2017). The sb and rb are estimated from the surface meteorological observations using the Monin–Obukhov flux–gradient similarity theory. For se and re, we assume that the air masses imported to the BL due to entrainment have properties of air in the lowermost free troposphere between 750 and 850 m MSL, which were obtained from the upper-air sounding data. To estimate scd and rcd, we assume that the convective downdraft air masses originate from the midtroposphere between 2500 and 3500 m MSL and are saturated due to evaporation of falling rainwater until they get to the surface. As MSE is conserved during this descent process, hcd equals to the midtropospheric MSE, which can be obtained from the upper-air sounding data. As the convective downdraft air mass is assumed to be saturated, hcd = scd + Lr*(scd/Cp), where r*(scd/Cp) is saturated mixing ratio at temperature scd/Cp calculated from the Clausius–Clapeyron equation. Since this equation implies that hcd is a monotonic increasing function of scd (note that saturated mixing ratio is a monotonic increasing function of temperature), we can calculate scd from this equation, and rcd = r*(scd/Cp). An analysis of the upper-air sounding data shows that the midtropospheric MSE, and thus hcd, scd, and rcd, are primarily determined by the midtropospheric humidity.
While the temporal resolution of the surface meteorological parameters is 10 min and that of the upper-air sounding data is 3 h, we use their daily means to calculate daily me and mcd. This is because the typical time scale for BLQE to become established over warm oceans can be estimated at approximately half a day (Raymond 1995; Raymond et al. 2015) and because the spatial scale in question is large enough to contain an ensemble of convective cells, whereas the data analyzed are point observations. Note that de Szoeke (2018) did not apply such temporal averages, which is another difference between de Szoeke (2018) and the present study.
In contrast to that of me and mcd, the sign of mu is defined as positive upward. The sign of mLS is defined as positive when there is horizontal convergence. We can obtain mu if we know mLS in addition to me and mcd estimated from Eqs. (1) and (2). As estimating mLS around the vessel is difficult, we instead use area-averaged horizontal convergence estimated from the upper-air observations at the three sites.
We assess statistical significance using parametric and two-sided tests assuming that each of the 22 days of the observation period is independent of the others and employing the 99% confidence level. For correlation analyses of daily time series, Student’s t test suggests that correlation coefficients greater than ±0.54 are statistically significant. The significance of a difference in averages between the preactive and active periods is assessed using Welch’s t test.
4. Results and discussion
a. Time series of coefficients in the budget equations
Figure 3 shows time series of daily mean quantities needed for estimating me and mcd. Time series of DSE (Fig. 3a) show that se (red dashed line) is higher than sb (black dotted line) and scd (blue solid line) is lower than sb throughout the observation period. This means that the entrainment contributes to the BL DSE import, whereas the convective downdraft contributes to its export. In contrast, both re and rcd are lower than rb (Fig. 3b) throughout the period, meaning that both processes contribute to the BL moisture export. These characteristics are typical over the tropical ocean with high SSTs. Among the three kinds of air masses, the convective downdraft air mass exhibits the largest variance in both DSE and mixing ratio. In addition to the day-to-day variability, there are increasing trends of scd and rcd over the course of the period. As one can infer from the way to estimate them explained in the last section, these increasing trends imply moistening trend in the midtroposphere. In fact, the upper-air observations showed gradual midtropospheric moistening over the course of the observation period (Katsumata et al. 2013). A slight decreasing trend of sb can also be observed, which is probably associated with the arrival of the convectively active phase of BSISO. These trends lead to decreasing trends of scd − sb and rcd − rb, whose influence on the mass fluxes will be discussed later.
Daily time series of (a) DSE divided by Cp and (b) mixing ratio of BL (black dotted line), entrainment (red dashed line), and convective downdraft (blue solid line) air masses, and daily time series of (c) SHF (purple solid line) and the radiative heating (orange dashed line) and (d) LHF.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
Figure 3c shows time series of SHF and radiative heating. There is an increasing trend in SHF (purple solid line), which is mainly due to a decreasing trend in surface air temperature indicated by the sb time series. Three maxima in SHF on 14, 21, and 26 June correspond to the maxima in the convective activity (Fig. 2a), which is likely related to convective enhancement (Young et al. 1995; Saxen and Rutledge 1998; Yokoi et al. 2014). Note that SHF is close to zero and sometimes negative in the preactive period. The average SHF over the preactive and active periods are +0.2 and +5.5 W m−2, respectively. In contrast to SHF, the radiative heating (orange dashed line) is always negative and less variable, with a 22-day average of −6.8 W m−2. As a result, H + R is generally negative except for 20, 21, and 26 June, with averages over the preactive and active periods of −7.2 and −0.8 W m−2, respectively. In contrast, as shown in Fig. 3d, LHF is always positive and ranges from 60 to 130 W m−2. Therefore, at least in this period, the entrainment and convective downdraft processes must jointly contribute to the BL DSE import and moisture export.
b. Day-to-day variability in estimated mass fluxes
Daily time series of estimated me and mcd are shown in Fig. 4a. The me ranges from 5 to 12 g s−1 m−2, with a slightly decreasing trend. In contrast, mcd exhibits an increasing trend with remarkable contrast between the preactive and active periods. Whereas mcd is less than 1 g s−1 m−2 in most of the preactive period, it ranges from 1 to 6 g s−1 m−2 in the active period. Although our method allows me and mcd to be negative, all of the estimated values are positive, suggesting that the estimation method works well. Furthermore, while the ratio of me to mcd depends on the heights where entrainment and convective downdraft air masses are assumed to originate, me is always larger than mcd in realistic ranges of the heights. The temporal variability of me and mcd is also qualitatively insensitive to the choice of the heights.
(a) Daily time series of me (red dashed line) and mcd (blue solid line). (b) Scatterplot of me (red triangles) and mcd (blue circles) against REC. (c) Daily time series of mu (solid line), me + mcd (dotted line), and mLS (dashed line). (d) Scatterplot of mu against LCC.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
The magnitudes of me and mcd are on the order of 1–10 g s−1 m−2. They are at least one order of magnitude larger than the downward mass flux that explains the observed tendency of zb (0.14 g s−1 m−2; see section 3). This means that disregarding tendency of zb in Eq. (4) is a good approximation.
The estimated mass fluxes are shown to have tight relationships with convective activity around the vessel. A comparison between the time series of mcd and REC, which is the fractional area with the radar echo stronger than 15 dBZ, indicates that the three maxima in REC (14, 21, and 26 June) are accompanied by maxima in mcd. A scatterplot of the mass fluxes against REC (Fig. 4b) reveals that mcd exhibits a positive and statistically significant correlation with REC. Note that the correlation is still statistically significant even if the day with exceptionally high REC (21 June) is excluded. Furthermore, mcd is close to zero for days with REC ∼0. The tight relationship between the intensity of convective downdrafts and the convective activity has been pointed out in previous studies (e.g., Yokoi et al. 2014). Note that, when we estimate the mass fluxes with the radiative heating neglected, mcd increases by 0.5–1 g s−1 m−2 and thus does not approach zero even for days with REC ∼ 0. This result suggests that considering the radiative heating is important for a physically consistent estimation of the mass fluxes. Note that some theorical studies also considered the radiative heating in the BL in developing the conceptual models (e.g., Raymond 1994, 1995).
The convective updraft mass flux (mu) can be estimated from me, mcd, and large-scale horizontal convergence (mLS) using Eq. (4). Figure 4c shows that me + mcd (dotted line) is generally larger than mLS (dashed line) and in between 6 and 13 g s−1 m−2. The mLS is positive in most of the preactive period, while it is negative in the first half of the active period (18–23 June), which is presumably associated with a large fraction of stratiform precipitation in the convectively active phase of BSISO (Xu and Rutledge 2018). The resultant mu (solid line) is positive throughout the period, ranging from 4 to 18 g s−1 m−2. Because the liquid condensation level is generally close to the BL top during the period (Geng et al. 2013), most of the updrafts across the BL top are expected to accompany condensation near the BL top. Furthermore, mu is considered to depend primarily on the fractional area of the updrafts rather than the mean updraft speed (Kumar et al. 2015). Therefore, variability in mu is expected to be associated with LCC. In fact, there are several correspondences between time series of mu and LCC. For example, both quantities are relatively small just after the arrival of the convectively active phase of BSISO. The scatterplot of mu against LCC (Fig. 4d) shows a positive and statistically significant correlation between them, suggesting that the estimated mass fluxes are also consistent with LCC. Note that the correlation coefficient between me + mcd and LCC is also statistically significant but lower than that between mu and LCC, which implies significant contribution of the large-scale horizontal convergence to mu.
The ratio of mcd to mu (mcd/mu) is considered a function of both cloud microphysics and large-scale thermodynamic conditions in the free troposphere (Raymond 1995; Emanuel 2019). Raymond (1995) speculated that mcd/mu is larger when the midtroposphere is drier. On the other hand, Emanuel (2019) argued that this ratio should increase with relative humidity in the free troposphere in radiative–convective equilibrium states. Figure 5 shows a scatterplot of this ratio against relative humidity in the midtroposphere where the convective downdraft air masses are assumed to originate. We show that larger mcd/mu is associated with wetter midtropospheric conditions, with a statistically significant correlation coefficient of +0.80.
Scatterplot of mcd/mu against midtropospheric relative humidity between 2500 and 3500 m MSL.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
The estimated downward mass fluxes have tight relationships with surface turbulent fluxes, as shown in Figs. 6a and 6b. Whereas mcd has no significant correlation with LHF, a significantly positive correlation between mcd and SHF can be observed. This relationship is probably associated with the enhancement of the surface fluxes by spreads of cold outflow after the convective downdraft reaches the surface. Many previous studies (e.g., Young et al. 1995; Jabouille et al. 1996) reported that such convective enhancement is more effective for SHF than LHF, consistent with the present results. In contrast to mcd, me has a significantly positive correlation with LHF. As LHF variability on relatively short time scales such as a daily scale is dominated by surface wind speed variability (DeMott et al. 2015; Yokoi 2020), me is expected to be dependent on the surface wind speed as well. Actually, me exhibits a significantly positive correlation with the surface wind speed (Fig. 6d).
Scatterplot of me (red triangles) and mcd (blue circles) against (a) SHF, (b) LHF, (c) hcd/Cp, and (d) surface wind speed.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
Figure 6c shows a scatterplot of the mass fluxes against hcd, latter of which is primarily determined by the midtropospheric humidity as explained in section 3. As demonstrated earlier, scd and rcd exhibit larger variance than the counterparts of the entrainment and BL air masses, suggesting that hcd also exhibits larger variance than hb. It thus seems interesting to examine the relationship between hcd and the mass fluxes. The scatterplot clearly shows a tight relationship between mcd and hcd. Note that this relationship is not linear but mcd seems to be an exponential function of hcd, with mcd increasing abruptly when hcd/Cp reaches ∼335 K. As mcd has tight relationship with convective activity and hcd is primarily determined by midtropospheric humidity, this exponential relationship is consistent with well-known relationships between column-integrated water vapor and precipitation (Bretherton et al. 2004; Peters and Neelin 2006) and between free-tropospheric humidity and precipitation (Holloway and Neelin 2009).
There are two mechanisms that can explain the relationship between mcd and hcd. It is known that, while entrainment of ambient free-tropospheric air into convective updrafts reduces their buoyancy and limit convective intensity, a moist environmental condition in the lower part of the free troposphere, in which estimated hcd is higher, diminishes this adverse effect on the convective systems (e.g., Brown and Zhang 1997; Holloway and Neelin 2009), leading to stronger convection and larger mcd. Furthermore, higher hcd leads to smaller difference between hcd and hb (note that hcd < hb). In such a case, as the contribution of the convective downdraft process to the BL MSE budget is mcd(hcd − hb), larger mcd is needed for this process to export required amount of MSE that is determined by the other three processes [see Eq. (3)]. In other words, efficiency of this process exporting BL MSE per unit mcd decreases with hcd. Note that hcd is assumed to be determined by the midtropospheric condition in this study and thus independent of mcd, suggesting a causal relationship that higher hcd contributes to larger mcd.
c. BL MSE budgets in different BSISO phases
In this subsection, we examine how the BL MSE budget processes change with the arrival of the convectively active phase of BSISO. First, the downward mass fluxes averaged over the preactive and active periods are compared in Table 1. While a difference in me between the two periods is small and insignificant, mcd increases more than 5 times in the active period compared with that in the preactive period, and the difference is statistically significant. It is likely that large portion of the increase in mcd is attributable to the increase in the midtropospheric humidity in the active period (Katsumata et al. 2013). As mentioned in the last subsection, there are two mechanisms how midtropospheric humidity affects mcd. Among them, here we estimate to what extent the efficiency of the convective downdraft air in exporting BL MSE explains the difference between the two periods. For this purpose, we substitute scd and rcd averaged over the 22-day period for daily scd and rcd in Eqs. (1) and (2) to calculate mass fluxes under fixed hcd condition. The fourth column of Table 1 shows the mcd thus calculated averaged over the two periods. The difference between the two periods decreases by approximately 26% compared with that in the original mcd, although it is still statistically significant. This suggests that approximately 1/4 of the mcd increase in the convectively active phase of BSISO is due to the decrease of the efficiency of the convective downdraft air in exporting BL MSE due to higher hcd and thus higher midtropospheric humidity.
The me, mcd, and mcd under the fixed hcd condition, averaged over the preactive and active periods, and their differences (active minus preactive) (unit: g s−1 m−2). Statistically significant differences are in boldface.
The differences in the mass fluxes result in those in the BL MSE budget. The BL MSE budget terms [see Eq. (3)] averaged over the two periods are shown in Fig. 7a. Positive values represent MSE import to the BL. In both periods, the surface turbulent flux imports the MSE, whereas the other three processes export it. The MSE import by the surface flux increases slightly in the active period compared with the preactive period, although the difference is statistically insignificant. As for the MSE export, the entrainment plays the largest role, and the convective downdraft plays the second largest role. The MSE exports by the entrainment and radiative process decrease in the active period. In contrast, the MSE export by the convective downdraft increases 3 times, with the difference statistically significant, which is primarily due to the increase in mcd.
Budgets of the BL (a) MSE, (b) DSE, and (c) mixing ratio multiplied by L averaged over (left) the preactive period and (center) the active period, and (right) the difference between the two periods (active minus preactive). Purple, red, blue, and orange bars represent the contributions of the surface flux, entrainment, convective downdraft, and radiative process, respectively. In the right panels, bars are enclosed by black lines if the difference is statistically significant at the 99% confidence level.
Citation: Journal of the Atmospheric Sciences 79, 3; 10.1175/JAS-D-21-0109.1
The budgets in the BL MSE can be decomposed into those in the DSE and moisture, and characteristics in the differences in the DSE and moisture budgets between the two periods (Figs. 7b,c) are generally similar to those in the MSE budget. The DSE import is caused by the surface flux and entrainment, with the latter playing a larger role, while its export is caused by the convective downdraft and radiative process. An increase in the DSE export by the convective downdraft in the active period is balanced with increases in the DSE import by the surface flux and entrainment and a decrease in the DSE export by the radiative process. The differences are statistically significant for all four processes. As for the moisture budget, the entrainment plays the dominant role in its export to counter its import by the surface flux. Differences between the two periods show that the moisture export by the convective downdraft increases significantly in the active period, while that by the entrainment decreases slightly.
As introduced in section 1, de Szoeke (2018) estimated the BL MSE exports by the convective downdraft and entrainment over the equatorial Indian Ocean in the suppressed, disturbed, and active phases of MJO in October–December 2011. Differences in the BL MSE budget between the suppressed and active phases of MJO revealed in de Szoeke (2018) have several similarities to those between the preactive and active periods in the present study. For example, de Szoeke (2018) showed that the MSE export by the convective downdraft in the active phase of MJO was ∼1.8 times larger than that in the suppressed phase, while the MSE export by the entrainment in the active phase was smaller than that in the suppressed phase.
On the other hand, relative importance of the entrainment and convective downdraft on the MSE export is different between the two studies. Results of de Szoeke (2018) reveal that, whereas both processes play essential roles, the convective downdraft tends to export more MSE than the entrainment, especially in the active phase of MJO when a ratio of the MSE export by the convective downdraft to that by the entrainment, mcd(hcd − hb)/me(he − hb), is 2.35. In contrast, our estimation suggests that the entrainment exports more MSE in both periods, with the ratio in the active period of 0.64 (e.g., Fig. 7a). The difference in the relative importance of the two processes between the two studies is partly due to the treatment of the radiative process. This study does not neglect the radiative process, whereas de Szoeke (2018) did so. As explained earlier, neglecting the radiative process leads to an overestimation of mcd. If we also neglect the radiative process, the estimated MSE export by the convective downdraft in the active period increases 1.3 times, and the ratio becomes 0.91. Other possible reasons for the difference include differences in the area and season of the observation and in the characteristics of large-scale disturbances that emerged during the observation period, implying the necessity of further research on the regionality and variability of the contributions of the two processes.
The radiative process appears to play only a secondary role in the change of the MSE and DSE budgets with BSISO (Figs. 7a,b); the difference in the radiative cooling in the BL between the two periods is only ∼1 W m−2. However, there is a possibility that we may underestimate the variability of the radiative heating. Shell et al. (2020) estimated vertical profile of the radiative heating over the equatorial Indian Ocean in October–December 2011, and their results indicate that the radiative cooling decreases 5–8 W m−2 in the active phase of MJO compared with the suppressed phase (their Fig. 12). As our estimation method may be too simple to examine variability of the radiative heating compared with that used in Shell et al. (2020), we need to sophisticate the method, which is one of the subjects of our future study.
5. Summary and conclusions
The boundary layer (BL) moist static energy (MSE) is considered one of the thermodynamic parameters that determine the intensity of cumulus convective activity in tropical oceanic regions with high SSTs. Discussions on the BL MSE budget by previous studies suggest that there are two important processes, the entrainment and convective downdraft, that cause downward mass fluxes across the BL top through which relatively low MSE air masses enter the BL, decreasing BL MSE per unit mass. This study estimated daily mean mass fluxes due to these two processes separately by analyzing upper-air sounding data and surface meteorological parameters observed over the tropical western Pacific in 6–27 June 2008 using R/V Mirai as the part of the PALAU2008 field campaign. The estimation method adopted employs budget equations of BL dry static energy (DSE) and water vapor mixing ratio, which is based on the method in de Szoeke (2018) with several modifications. Results shown in this paper demonstrate that estimations of the convective downdraft mass flux (mcd) and entrainment mass flux (me) are physically reasonable and thus is expected to be helpful in understanding the interactions between convective activity and atmospheric large-scale circulation fields over tropical oceanic areas.
It was shown that the estimated mass fluxes were consistent with convective activity. There is a statistically significant correlation between mcd and the fractional area with radar echoes (REC) around the vessel, and mcd is close to zero for days with almost no echoes. Besides, the updraft mass flux (mu), which is the sum of mcd, me, and large-scale BL horizontal convergence (mLS), correlates significantly with fractional time with low-level cloud cover (LCC).
The mass fluxes also exhibit close relationships with surface turbulent fluxes. In particular, mcd correlates significantly with sensible heat flux (SHF), which is likely associated with the enhancement of surface fluxes due to convective cold outflow. There is also significant correlation between me and latent heat flux (LHF). We also revealed exponential relationship between mcd and midtropospheric humidity, which is consistent with findings by previous studies (Bretherton et al. 2004; Peters and Neelin 2006; Holloway and Neelin 2009).
Using the estimated mass fluxes, we showed remarkable differences in how the BL MSE budget is accomplished in different phases of the boreal summer intraseasonal oscillation (BSISO). During the study period, the convectively active phase of BSISO migrated northward and arrived at the observation area in the latter half of the period. The comparison of the mass fluxes and budget terms between the former half (preactive period) and latter half (active period) of the 22-day period revealed that mcd increased 5 times in the active period compared with that in the preactive period. Approximately 1/4 of the increase is caused by the decrease in the efficiency of the convective downdraft process in exporting BL MSE per unit mcd, which is caused by higher hcd and higher midtropospheric humidity. As a result of the increase in mcd, the BL MSE export by the convective downdraft increases 3 times in the active period, which is balanced by the increase in the MSE import by the surface flux and the decreases in the MSE export by the entrainment and radiative process. Differences in the BL DSE and moisture budgets between the two periods have characteristics similar to those in the MSE budget. The results on the differences in the BL MSE budget between the preactive and active periods are consistent with those between the suppressed and active phases of MJO over the equatorial Indian Ocean in boreal winter season examined by de Szoeke (2018); for example, the MSE export by the convective downdraft in the active phase of MJO increases ∼1.8 times compared with the suppressed phase.
Our results showed that the MSE export by the entrainment was larger than that by the convective downdraft in both periods. This is different from the results of de Szoeke (2018), which concluded that the convective downdraft tends to export more MSE than the entrainment. The difference partly arises from considering the radiative heating for the BL DSE budget in the present study. It is also possible that the relative importance of the convective downdraft to the entrainment is regionally and seasonally dependent. Thus, it seems interesting to analyze shipborne observations obtained in other field campaigns in the Indo-Pacific warm pool domain and long-term station observations in tropical islands to accumulate information on the characteristics of the BL MSE budget, which we will pursue in the future.
Acknowledgments.
The authors are grateful to all those who engaged in the R/V Mirai MR08-02 cruise. We also thank Prof. Simon P. de Szoeke and one anonymous reviewer for improving this manuscript. This study is partly supported by Grant-in-Aid for Scientific Research (16H04048, 20H01386, 20H02252) of the Japan Society for the Promotion of Science.
Data availability statement.
The R/V Mirai MR08-02 observational data are openly available from JAMSTEC DARWIN database at http://www.godac.jamstec.go.jp/darwin/e. The GSMaP dataset is openly available via https://sharaku.eorc.jaxa.jp/GSMaP/index.htm. The JRA55 dataset is openly available via https://jra.kishou.go.jp/JRA-55/index_en.html.
APPENDIX
Estimation of the Radiative Heating
Here,
Here
In this study, we decide to set ϵ = 0.14 and η = 0.02 so that the daily mean R will be close to −1 K day−1 and the difference between daily maximum and minimum of R will be close to +3 K day−1, using examples from estimated radiative heating rate over the equatorial Indian Ocean in October–December 2011 (de Szoeke et al. 2017).
REFERENCES
Ahmed, F., and J. D. Neelin, 2018: Reverse engineering the tropical precipitation–buoyancy relationship. J. Atmos. Sci., 75, 1587–1608, https://doi.org/10.1175/JAS-D-17-0333.1.
Arakawa, A., and W. H. Schubert, 1974: Interaction of a cumulus cloud ensemble with the large-scale environment, Part I. J. Atmos. Sci., 31, 674–701, https://doi.org/10.1175/1520-0469(1974)031<0674:IOACCE>2.0.CO;2.
Bretherton, C. S., M. E. Peters, and L. E. Back, 2004: Relationships between water vapor path and precipitation over the tropical oceans. J. Climate, 17, 1517–1528, https://doi.org/10.1175/1520-0442(2004)017<1517:RBWVPA>2.0.CO;2.
Brown, R. G., and C. Zhang, 1997: Variability of midtropospheric moisture and its effect on cloud-top height distribution during TOGA COARE. J. Atmos. Sci., 54, 2760–2774, https://doi.org/10.1175/1520-0469(1997)054<2760:VOMMAI>2.0.CO;2.
Cheng, M.-D., and A. Arakawa, 1997: Inclusion of rainwater budget and convective downdrafts in the Arakawa-Schubert cumulus parameterization. J. Atmos. Sci., 54, 1359–1378, https://doi.org/10.1175/1520-0469(1997)054<1359:IORBAC>2.0.CO;2.
DeMott, C. A., N. P. Klingaman, and S. J. Woolnough, 2015: Atmosphere-ocean coupled processes in the Madden-Julian oscillation. Rev. Geophys., 53, 1099–1154, https://doi.org/10.1002/2014RG000478.
de Szoeke, S. P., 2018: Variations of the moist static energy budget of the tropical Indian Ocean atmospheric boundary layer. J. Atmos. Sci., 75, 1545–1551, https://doi.org/10.1175/JAS-D-17-0345.1.
de Szoeke, S. P., E. D. Skyllingstad, P. Zuidema, and A. S. Chandra, 2017: Cold pools and their influence on the tropical marine boundary layer. J. Atmos. Sci., 74, 1149–1168, https://doi.org/10.1175/JAS-D-16-0264.1.
Emanuel, K. A., 1989: The finite-amplitude nature of tropical cyclogenesis. J. Atmos. Sci., 46, 3431–3456, https://doi.org/10.1175/1520-0469(1989)046<3431:TFANOT>2.0.CO;2.
Emanuel, K. A., 2019: Inferences from simple models of slow, convectively coupled processes. J. Atmos. Sci., 76, 195–208, https://doi.org/10.1175/JAS-D-18-0090.1.
Fairall, C. W., E. F. Bradley, J. E. Hare, A. A. Grachev, and J. B. Edson, 2003: Bulk parameterization of air–sea fluxes: Updates and verification for the COARE algorithm. J. Climate, 16, 571–591, https://doi.org/10.1175/1520-0442(2003)016<0571:BPOASF>2.0.CO;2.
Garstang, M., and A. K. Betts, 1974: A review of the tropical boundary layer and cumulus convection: Structure, parameterization, and modeling. Bull. Amer. Meteor. Soc., 55, 1195–1205, https://doi.org/10.1175/1520-0477(1974)055<1195:AROTTB>2.0.CO;2.
Geng, B., K. Yoneyama, R. Shirooka, and M. Yoshizaki, 2011: Characteristics of precipitation systems and their environment observed during the onset of the western North Pacific summer monsoon in 2008. J. Meteor. Soc. Japan, 89A, 1–25, https://doi.org/10.2151/jmsj.2011-A01.
Geng, B., K. Yoneyama, and R. Shirooka, 2013: Thermodynamic structure and evolution of the atmospheric mixed layer over the western North Pacific during the summer monsoon onset. J. Geophys. Res. Atmos., 118, 5655–5666, https://doi.org/10.1002/jgrd.50242.
Geng, B., K. Yoneyama, and R. Shirooka, 2014: Observations of upper-tropospheric influence on a monsoon trough over the western North Pacific. Mon. Wea. Rev., 142, 1472–1488, https://doi.org/10.1175/MWR-D-13-00233.1.
Holloway, C. E., and J. D. Neelin, 2009: Moisture vertical structure, column water vapor, and tropical deep convection. J. Atmos. Sci., 66, 1665–1683, https://doi.org/10.1175/2008JAS2806.1.
Jabouille, P., J. L. Redelsperger, and J. P. Lafore, 1996: Modification of surface fluxes by atmospheric convection in the TOGA COARE region. Mon. Wea. Rev., 124, 816–837, https://doi.org/10.1175/1520-0493(1996)124<0816:MOSFBA>2.0.CO;2.
JAMSTEC, 2008: R/V Mirai Cruise Rep. MR08-02, JAMSTEC, 126 pp., http://www.jamstec.go.jp/dcop/e/publications/pdf/CruiseReport_MR08-02_all.pdf.
Katsumata, M., H. Yamada, H. Kubota, Q. Moteki, and R. Shirooka, 2013: Observed evolution of northward-propagating intraseasonal variation over the western Pacific: A case study in boreal early summer. Mon. Wea. Rev., 141, 690–706, https://doi.org/10.1175/MWR-D-12-00011.1.
Kikuchi, K., 2021: The boreal summer intraseasonal oscillation: A review. J. Meteor. Soc. Japan, 99, 933–972, https://doi.org/10.2151/jmsj.2021-045.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Kubota, T., and Coauthors, 2020: Global Satellite Mapping of Precipitation (GSMaP) products in the GPM era. Satellite Precipitation Measurement, V. Levizzani et al., Eds., Advances in Global Change Research, Vol. 67, Springer, 355–373.
Kumar, V. V., C. Jakob, A. Protat, C. R. Williams, and P. T. May, 2015: Mass-flux characteristics of tropical cumulus clouds from wind profiler observations at Darwin, Australia. J. Atmos. Sci., 72, 1837–1855, https://doi.org/10.1175/JAS-D-14-0259.1.
Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708, https://doi.org/10.1175/1520-0469(1971)028<0702:DOADOI>2.0.CO;2.
Madden, R. A., and P. R. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29, 1109–1123, https://doi.org/10.1175/1520-0469(1972)029<1109:DOGSCC>2.0.CO;2.
Peters, O., and J. D. Neelin, 2006: Critical phenomena in atmospheric precipitation. Nat. Phys., 2, 393–396, https://doi.org/10.1038/nphys314.
Raymond, D. J., 1994: Convective processes and tropical atmospheric circulations. Quart. J. Roy. Meteor. Soc., 120, 1431–1455, https://doi.org/10.1002/qj.49712052002.
Raymond, D. J., 1995: Regulation of moist convection over the west Pacific warm pool. J. Atmos. Sci., 52, 3945–3959, https://doi.org/10.1175/1520-0469(1995)052<3945:ROMCOT>2.0.CO;2.
Raymond, D. J., Ž. Fuchs, S. Gjorgjievska, and S. Sessions, 2015: Balanced dynamics and convection in the tropical troposphere. J. Adv. Model. Earth Syst., 7, 1093–1116, https://doi.org/10.1002/2015MS000467.
Saxen, T. R., and S. A. Rutledge, 1998: Surface fluxes and boundary layer recovery in TOGA COARE: Sensitivity to convective organization. J. Atmos. Sci., 55, 2763–2781, https://doi.org/10.1175/1520-0469(1998)055<2763:SFABLR>2.0.CO;2.
Shell, K. M., S. P. de Szoeke, and M. Makiyama, 2020: Vertical structure of radiative heating rates of the MJO during DYNAMO. J. Climate, 33, 5317–5335, https://doi.org/10.1175/JCLI-D-19-0519.1.
Thayer-Calder, K., and D. Randall, 2015: A numerical investigation of boundary layer quasi-equilibrium. Geophys. Res. Lett., 42, 550–556, https://doi.org/10.1002/2014GL062649.
Xu, W., and S. A. Rutledge, 2018: Convective variability associated with the boreal summer intraseasonal oscillation in the South China Sea region. J. Climate, 31, 7363–7383, https://doi.org/10.1175/JCLI-D-18-0091.1.
Yanai, M., S. Esbensen, and J.-H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci., 30, 611–627, https://doi.org/10.1175/1520-0469(1973)030<0611:DOBPOT>2.0.CO;2.
Yokoi, S., 2020: Diurnal variation of surface heat fluxes off the west coast of Sumatra Island as revealed by in situ observation. SOLA, 16A, 1–5, https://doi.org/10.2151/sola.16A-001.
Yokoi, S., M. Katsumata, and K. Yoneyama, 2014: Variability in surface meteorology and air-sea fluxes due to cumulus convective systems observed during CINDY/DYNAMO. J. Geophys. Res. Atmos., 119, 2064–2078, https://doi.org/10.1002/2013JD020621.
Yoneyama, K., C. Zhang, and C. N. Long, 2013: Tracking pulses of the Madden–Julian oscillation. Bull. Amer. Meteor. Soc., 94, 1871–1891, https://doi.org/10.1175/BAMS-D-12-00157.1.
Young, G. S., S. M. Perugini, and C. W. Fairall, 1995: Convective wakes in the equatorial western Pacific during TOGA. Mon. Wea. Rev., 123, 110–123, https://doi.org/10.1175/1520-0493(1995)123<0110:CWITEW>2.0.CO;2.
