Convective cloud development during the Indian monsoon helps moisten the atmospheric environment and drive the monsoon trough northwards each year, bringing a large amount of India's annual rainfall. Therefore, an increased understanding of how monsoon convection develops from observations will help inform model development. In this study, 139 days of India Meteorological Department Doppler weather radar data is analyzed for seven sites across India during the 2016 monsoon season. Convective cell‐top heights (CTHs) are objectively identified through the season, and compared with near‐surface (at 2 km height) reflectivity. These variables are analyzed over three time scales of variability during the monsoon: monsoon progression on a month‐by‐month basis, active‐break periods and the diurnal cycle. We find a modal maximum in CTH around 6–8 km for all sites. Cell‐averaged reflectivity increases with CTH, at first sharply, then less sharply above the freezing level. Bhopal and Mumbai exhibit lower CTH for monsoon break periods compared to active periods. A clear diurnal cycle in CTH is seen at all sites except Mumbai. For south‐eastern India, the phase of the diurnal cycle depends on whether the surface is land or ocean, with the frequency of oceanic cells typically exhibiting an earlier morning peak compared to land, consistent with the diurnal cycle of precipitation. Our findings confirm that Indian monsoon convective regimes are partly regulated by the large‐scale synoptic environment within which they are embedded. This demonstrates the excellent potential for weather radars to improve understanding of convection in tropical regions.

<p>Convective cloud development during the Indian monsoon helps moisten the atmospheric environment and drive the monsoon trough northwards each year, bringing a large amount of India’s annual rainfall. Therefore, an increased understanding of how monsoon convection develops in observations will help inform model development. In this study, 139 days of India Meteorological Department Doppler weather radar data is analysed for 7 sites across India during the 2016 monsoon season. Convective cell-top heights (CTH) are objectively identified through the season, and compared with near-surface (at 2 km height) reflectivity. These variables are analysed over three time scales of variability during the monsoon: monsoon progression, active-break periods and the diurnal cycle. We find a modal maximum in CTH around 6–8 km for all sites. Reflectivity increases with CTH, at first sharply, then less sharply above the freezing level. Bhopal and Mumbai exhibit lower CTH for monsoon break periods compared to active periods. A clear diurnal cycle in CTH is seen at all sites except Mumbai. The phase of the diurnal cycle depends on the surface type being land or ocean for south-eastern India, with the frequency of oceanic cells typically exhibiting an early morning peak compared to those over land, consistent with the observed diurnal cycle of precipitation. The cell characteristics discovered are discussed in light of the differences in large-scale synoptic and mesoscale mechanisms responsible for different cell regimes. Our findings confirm that Indian monsoon convective regimes are partly regulated by the large-scale synoptic environment within which they are embedded.</p>

Figure 3. 5-day running means (centred on the date shown) of local precipitation from IMD 1° × 1° gridded daily precipitation (spatially averaged for the nearest five grid boxes). Also shown is the daily sum of deep (CTH ≥ 8km) and congestus (5 ≤ CTH < 8km) convective area for each day (normalised by the number of radar volume scans) at each site for when data are available. Pearson product-moment correlation coefficients of precipitation against deep and congestus cell area over the time series are labelled for each site in the top right corner. Grey shaded regions are days with 0 radar volume scans for that site and are not included in the correlation. The vertical black dashed line represents the day of onset for that site, as given in Table 1.

Convective cloud development during the Indian monsoon helps moisten the atmospheric environment and drive the monsoon trough northwards each year, bringing a large amount of India's annual rainfall. Therefore, an increased understanding of how monsoon convection develops from observations will help inform model development. In this study, 139 days of India Meteorological Department Doppler weather radar data is analyzed for seven sites across India during the 2016 monsoon season. Convective cell-top heights (CTHs) are objectively identified through the season, and compared with near-surface (at 2 km height) reflectivity. These variables are analyzed over three time scales of variability during the monsoon: monsoon progression on a month-by-month basis, active-break periods and the diurnal cycle. We find a modal maximum in CTH around 6-8 km for all sites. Cell-averaged reflectivity increases with CTH, at first sharply, then less sharply above the freezing level. Bhopal and Mumbai exhibit lower CTH for monsoon break periods compared to active periods. A clear diurnal cycle in CTH is seen at all sites except Mumbai. For south-eastern India, the phase of the diurnal cycle depends on whether the surface is land or ocean, with the frequency of oceanic cells typically exhibiting an earlier morning peak compared to land, consistent with the diurnal cycle of precipitation. Our findings confirm that Indian monsoon convective regimes are partly regulated by the large-scale synoptic environment within which they are embedded. This demonstrates the excellent potential for weather radars to improve understanding of convection in tropical regions.

&lt;p&gt;Operational ground-based weather radar data provide a unique opportunity to evaluate simulations of convection in high-resolution models. This is especially advantageous for tropical regions such as India where convection is frequent and interacts with the large-scale circulation. Here, storms derived from 15 Indian operational radars&amp;#160;are directly compared to modelled storms at two convection-permitting resolutions, 1.5 and 4.4 km, for a period of 3 weeks during the peak 2016 monsoon season. We objectively identify different morphological properties of storms for 6 regions of India, that is to say their heights, sizes, and intensities. Both model resolutions are found to simulate too much shallow convection compared to radars for all 6 regions. Modelled convection is also frequently too wide and intense, but the 1.5 km model performs noticeably better. Modelled storms also exhibit a maximum area around the freezing level, higher than observed by radars, especially at 4.4 km resolution. We discuss various potential microphysical and dynamical reasons for the major differences seen, thus demonstrating the power of radar-based evaluation of monsoon convection for the Indian region.&lt;/p&gt;