Convective environment in pre-monsoon and monsoon conditions over the Indian subcontinent: the impact of surface forcing

Thermodynamic soundings for pre-monsoon and monsoon seasons from the Indian subcontinent are analysed to document differences between convective environments. The pre-monsoon environment features more variability for both near-surface moisture and free-tropospheric temperature and moisture profiles. As a result, the level of neutral buoyancy (LNB) and pseudo-adiabatic convective available potential energy (CAPE) vary more for the pre-monsoon environment. Pre-monsoon soundings also feature higher lifting condensation levels (LCLs). LCL heights are shown to depend on the availability of surface moisture, with low LCLs corresponding to high surface humidity, arguably because of the availability of soil moisture. A simple theoretical argument is developed and showed to mimic the observed relationship between LCLs and surface moisture. We argue that the key element is the partitioning of surface energy flux into its sensible and latent components, that is, the surface Bowen ratio, and the way the Bowen ratio affects surface buoyancy flux. We support our argument with observations of changes in the Bowen ratio and LCL height around the monsoon onset, and with idealized simulations of cloud fields driven by surface heat fluxes with different Bowen ratios.

Note that CAPE is given as cCAP E(z = LN B). In addition, the equivalent potential temperature θ e was calculated using 106 approximate formula: where L is latent heat of condensation.  For pre-monsoon conditions, the surface temperature is on average several degrees warmer and water vapor mixing ratio is on 116 average about half of that for monsoon period. The latter is arguably related to the contrasting levels of soil moisture in pre-117 monsoon and monsoon conditions. The temperature and moisture profiles exhibit less day-to-day variability for the monsoon 118 period. The spread of temperature in middle troposphere in the monsoon environment is about half of that for pre-monsoon. 119 In upper troposphere and lower stratosphere, the differences are smaller. For the pre-monsoon period, moisture profiles below 120 Fig. 1. Profiles of potential temperature (left panels), water vapor mixing ratio (middle panels), and relative humidity (right panels) for pre-monsoon (upper row) and monsoon (lower row) soundings.
6 km vary significantly and atmosphere is significantly drier above 6 km when compared to monsoon soundings. Arguably, 121 higher moisture contents in the middle and upper troposphere during monsoon come from convection reaching higher levels 122 as documented later in the paper. However, differences due to large-scale horizontal advection may play some role as well.

123
Individual moisture profiles feature significant fluctuations, even more apparent if no smoothing is applied to the original 124 high resolution data. This is evident at lower levels (i.e., within the boundary layer) as well as aloft. Fluctuations within 125 boundary layer show that it is not well-mixed for the water vapor in most soundings, especially for monsoon conditions.

126
However, relative humidity does increase approximately linearly within the boundary layer in most profiles similar to the case 127 of well-mixed mixed boundary layer (i.e., featuring constant with height potential temperature and water vapor mixing ratio). moisture profile observations with microwave radiometer profiler. This is because the adiabatic (neutral) temperature profile 132 (i.e., constant θ) within the well-mixed boundary layer has to change to stably-stratified profile (i.e., θ increasing with height) 133 in the free troposphere aloft. Since LCL marks the transition from dry to moist temperature lapse rate within a rising adiabatic surface energy budget, but we neglect this aspect for the qualitative discussion here. Thus, we assume that development of 145 convective boundary layer during pre-monsoon and monsoon periods is to the leading order affected by partitioning of total 146 surface energy flux into its sensible and latent components, and not by the differences in total flux.

147
The partitioning of surface flux into sensible and latent components depends on the soil moisture that differs drastically 148 between pre-monsoon and monsoon conditions. The surface buoyancy flux that drives boundary layer dynamics is affected 149 by the surface Bowen ratio. Since the thermodynamic variable relevant for buoyancy flux is the virtual potential temperature θ o is the surface potential temperature.Total surface energy flux EF can be similarly written (using the moist static energy or 152 the equivalent potential temperature) as EF =< wθ > + L cp < wq v >. Consequently, BF/EF ratio between the buoyancy and 153 energy surface fluxes can then be represented as: L<wqv> is the Bowen ratio. For small Bowen ratios (i.e., surface 156 latent heat flux dominates as typically over the oceans) the BF/EF ratio approaches 0.1, that is, only 10 % of the total surface The above considerations explain the well-known fact that daytime convective boundary layer develops deep over arid and 163 semi-arid areas that feature high Bowen ratio due to limited availability of water at the surface. For instance, over the Sahara 164 desert, the boundary layer height can reach several kilometres (e.g., Ao et al., 2012). In contrast, surface-driven convective 165 boundary layer over tropical and sub-tropical oceans is relatively shallow, often a mere several hundred meters. We argue 166 that the differences between pre-monsoon and monsoon periods can, to a large degree, be explained by the availability of soil 167 moisture and partitioning of surface energy flux between sensible and latent components. These differences will be further 168 illustrated by model simulations discussed in section 5. to the north and easterlies to the south that are associated with mid tropospheric anticyclone.

179
In the case of pre-monsoon conditions, the moisture availability in BL is considerably reduced and this has a significant 180 influence on cloud base height. Air parcels need to rise to greater heights in pre-monsoon conditions to reach LCL compared to monsoon conditions. Significant variations are observed in LCL heights during these two seasons. Pre-monsoon clouds have 182 their bases at higher levels, 2 to 6 km from the surface, whereas monsoon soundings indicate cloud bases at lower levels with 183 most of them being lower than 2 km. This result is highly correlated with surface level moisture as documented below.

184
BL as well as mid-tropospheric moisture for the two seasons exhibit contrasting characteristics. The mean tropospheric 185 moisture is higher for monsoon soundings. During monsoon, the surface values of q v are higher compared to pre-monsoon, 186 and most of them fall within the range of 14-18 gkg −1 . Pre-monsoon surface q v has a lower but wider range from 3 to 14 187 gkg −1 . Monsoon soundings also indicate higher levels of mid-tropospheric moisture. The main reason is south westerly winds 188 that transport moisture from Arabian Sea to Indian subcontinent. Because of Western Ghat mountains, the transport features 189 strong low-level convergence over Indian west coast. However, for the inland locations over rain shadow region, the jet core 190 level is seen at 1.5-2 km, just above the boundary layer. Arguably, boundary layer convection developing during the day pushes 191 jet layer to an elevated height. For a well-mixed boundary layer, water vapor mixing ratio near surface is the main determining factor for cloud base height. gives: over the Indian subcontinent. Since the column PW is dominated by moisture in the lowest levels (and in the boundary layer in particular), the mixing ratio near the surface should then be well correlated with LCL height as documented in Fig. 3.

234
The above results can also be used in reverse. The fact that, despite some offset, there is an almost a perfect relationship 235 between RH and z LCL implies that mid-day boundary layer for all soundings considered in this study is of convective type, 236 that is, with close to the adiabatic potential temperature profile from above the superadiabatic surface layer up to the convective 237 boundary layer height and LCL. vection. Figure 4 shows profiles of pseudo-adiabatic buoyancy (i.e., the difference in the virtual potential temperature between 241 pseudo-adiabatic parcel and the environment) and cCAPE from all soundings separated into pre-monsoon and monsoon con- to monsoon environments and LNBs in the upper troposphere. One distinct feature of high-CAPE pre-monsoon category is that 261 the positive buoyancy increases steeply above LFC compared to the monsoon cases where buoyancy increased gradually above 262 the boundary layer. This is possibly due to the stark difference in moisture above LFC between pre-monsoon and monsoon 263 environments and its impact on the psuedo-adiabatic buoyancy.