Abstract
This study documents the climatological feature (1951–1980) and recent changes (1981–2020) in rainfall characteristics considering the observed nearly full spectrum of rain event sizes (daily contiguous rain area (CRA) events) in all seasons over India. It is found that the low frequency very large CRA (~synoptic scale) from monsoon season contributes ~50% of annual rainfall. However, the small-sized CRA (isolated thunderstorms) are the most frequent daily rain events (~70% of annual frequency) and hence are important for rain-fed agricultural practices. The well-documented widespread drying trend in the monsoon season has manifested in the annual rainfall trend but with reduced magnitude illustrating the compensatory effect from other seasons. Spatial aggregated annual statistics show that there is no significant change in rainfall amount and frequency of occurrence of rain events in the recent past compared to the base period. However, seasonally the pre-monsoon rainfall amount has increased significantly. Annually, the number of extremely heavy CRA (EHR) events have significantly increased by ~55% owing to a significant increase in pre-monsoon and monsoon rainfall. In all seasons, small-sized extremely heavy CRA has intensified substantially by 50–200% as compared to the base period. Additionally, the rain events from areal category large (~Mesoscale Convective Complexes (MCC)) have intensified in all seasons except winter. Thus, to decrease the uncertainty in rain-fed agricultural practices and better prediction of EHR to develop effective climate change mitigation strategies; process studies beyond monsoon season and processes other than synoptic scales are also required.
Similar content being viewed by others
Introduction
India is an agrarian country (~60% of the total population depends on agriculture), the agricultural practices span throughout the year and its major portion is rain-fed (~55% of the net sown area1,2); significantly affected by rainfall variability3,4,5. Seasonwise, the north and northwest parts of India receive a noticeable amount of rain in the winter season which is thus important for the rabi crops4,6. Pre-monsoon witnesses some of the severe local thunderstorms near the foothills of the Himalayas, eastern India, the southeast coast, and the extreme southwest coast7,8 that have the potential to cause a landslide and flash flood. On a similar note, the southeast part of India receives maximum rainfall during the northeast monsoon (post-monsoon) season5,9. Despite these imperative facts, the research studies to understand the rainfall characteristics, processes and trends in seasons other than the monsoon season have received considerably less attention10,11,12,13.
The summer monsoon being the prime rainfall contributing season to the annual rainfall, the mean14 and extreme15,16 characteristics of it, has been extensively studied in the past and has been reviewed17,18. Furthermore, the region of central India is the dominant region in receiving rain in the monsoon season and having relatively homogeneous variability in seasonal mean and daily rainfall15,19 (and references cited therein), has drawn much attention by the research community. These mentioned studies have found that frequency of extreme rainfall events have increased significantly over central India, however, mean summer rainfall does not show any appreciable trend.
Most of the extreme rainfall studies over India are grid point19,20,21 or station-level studies22,23 investigating frequency, intensity and temporal patterns of precipitation under changing climate. The grid point or station level studies can have a significant positive bias in projecting the frequency of extreme events as each grid point value is referred to as an independent extreme event however, in reality, they belong to a single weather system. Nevertheless, there are a few recent studies that have analyzed post-1950 rainfall extremes as a spatially collocated object but are limited to monsoon season24 and over central India15,25. From the point of view of water management through dams and flood forecasting, the information on rainfall intensity and areal extent24,26 both are important. Also, relating the spatial size of a rain event to the change in frequency and intensity provides a better understanding of the underlying atmospheric process by broadly categorizing the rain event into a weather system such as an isolated convective system, mesoscale convective complex or synoptic systems27. This study further suggests that spatial scales used to classify rain systems are more robust than using temporal scales.
Regardless of its diverse applicability, to the best of our knowledge, there is no study documenting changes in the rainfall characteristics by taking into account the full spectrum of rain event sizes and seasons over India. Thus, this study is a step forward to enhance our understanding of the characteristics of CRA (defined in next section) bridging over all seasons.
Results
Before going into the detailed study of the characteristics of CRA, some general rainfall features over Indian land are introspected. Figure 1 shows the seasonal and annual trend of precipitation over the Indian land at each grid point for the period 1951 to 2020. The non-parametric trends significant at a 95% confidence level are stippled28. The annual precipitation (Fig. 1e) over a majority of the Indian land has a decreasing trend with a statistically significant decrease over the Indo-Gangetic plain, central India, Western Ghats, and northeast India. On contrary, north-west India, western India, parts of the east coast and the interior of the Southern Peninsula are observed to have a positive annual rainfall trend. Further splitting the data into seasonal mean, peak rainfall contributor season monsoon (JJAS, Fig. 1c), shows a significant decreasing trend of rainfall over central, eastern, and northeast India as noted by many previous studies15,19,21 (and references mentioned therein). The majority of rainfall delivered during this season is from synoptic-level low-pressure systems.
The considered seasons are (a) winter (DJF), (b) pre-monsoon (MAM), (c) monsoon (JJAS), (d) post-monsoon (ON) and (e) Annual. Trends are computed using the Theil-Sen linear trend estimator. Changes significant at 95% confidence level using non-parametric Man-Kendall test are stippled. Trends are computed from yearly mean and then multiplied by 10 to represent decadal changes.
In the season winter (DJF, Fig. 1a), north and northwest India experiences predominantly non-convective weather activities and precipitation is advective in origin, associated with the passage of mid-latitude western disturbances (11 Roy Bhowmik et al, 2008). The resulting precipitation affects the winter crops sawed in northwest India. In this season an overall, weak negative trend is observed over north-east, western and central India. However, some interior parts of the southern peninsula are observed to have a significant positive trend. The pre-monsoon (MAM) season, is the hottest season and witnesses a significant number of weak to severe local thunderstorms under the influence of prevailing Nor’westerly winds and a convectively unstable environment7,11,29,30. In this season (Fig. 1b), the majority of the Indian land has a positive rainfall trend with a significant rise in the west, north-west and coastal east India. The post-monsoon season (ON, Fig. 1d) is also the season of northeast monsoon rainfall over the southeast coast and northeast India. In general, a weak negative rainfall trend is observed during the post-monsoon season except in northeast India, where it has a strong significant decreasing trend.
Areal distribution
To quantify the areal spread and seasonal cycle of daily CRA objects (here onwards referred to as rain event) box and whiskers diagram is shown in Supplementary Fig. 1. From here onwards the study is divided and compared within two periods; the base period (1951–1980) and the recent period (1981–2020). The choice of these periods is described in detail under the data and methodology section at the end of the article.
In all seasons, in the base period (blue box and whiskers), nearly 70–75% of the rain events have a size of less than 10,000 km2. In monsoon, 1% of the events can reach a size of more than 1,000,000 km2. Comparing the interquartile range of the areal distribution from the recent past to the base period, a negligible change (relative to the order of the magnitude) in the areal distribution is observed. Therefore, the areal category differing by an order of magnitude defined below are expected to be fairly independent of each other. Further, the classification of rain events into different areal categories and their correspondence to real weather systems are summarized in Table 1.
Table 1 has associated the areal categories defined in the present study with corresponding actual weather systems by making use of available past literature (citation mentioned in Table 1). Here it is important to mention that the horizontal scales used in the present study have comparable correspondence with the scales defined in the literature. As past length scale studies mostly used cloud-dependent parameters (cloud top temperature (brightness temperature), outgoing long wave, radar reflectivity etc.) but with some minimum threshold value to match the area of rainfall within the cloud system. Here, this study considered the minimum rainfall threshold as more than zero for all seasons except 1 mm for monsoon season to agree well with the threshold demarcating the weather system sizes described in the literature. The reason for the use of 1 mm in monsoon season is to avoid merging of weather systems due to interpolation of rain values. However, it is chosen to be low since multiple localized heavy rain areas could be a part of a single synoptic system. Given these facts, the physical interpretations drawn related to underlying physical processes related to weather systems in this study may have good correspondence with past literature. Here it is worth noticing that the mesoscale convective system (MCS) includes mesoscale convective complexes (MCC) but in this study, MCC are considered in a separate category and areal extent less the 100k but more than 10k are termed as MCS.
Frequency
Spatial distribution of the change in frequency of CRA
Figure 2 shows the change in the frequency of occurrence of CRA in the recent period compared to the base period. In all seasons small events are observed everywhere over the Indian land mass. The frequency of occurrence of these small events has significantly increased over Odisha (North-East of the eastern coast) and the interior of the Western Ghats in the recent period. This rainfall may be contributed from weak synoptic systems. A significant decrease in frequency is observed near the southern tip of the east coast in monsoon and post-monsoon seasons. The contribution of the very large events is mostly from the monsoon season over eastern India with a mix of positive and negative anomalies. The annual distribution shows a significant decrease in the occurrence of very large CRA in some regions of eastern India. Since Indian land is full of inhomogeneity in topography and climatic regimes along with the occurrence of different seasons, finding direct cause and effect is not simple. However, through analysis of spatially aggregated and seasonally differentiated studies along with the consideration of inter-annual variability, we are able to ascertain some more valuable conclusions.
The black stippling denotes a significant change (compared to inter-annual variability using 2-tailed t test) at 95% confidence level. For clarity, unit is represented in decadal change, however computed annually.
Spatially aggregated distribution of CRA frequency
Figure 3 shows the climatology of the base period and change observed in the recent past Fig. 3a depicts spatially aggregated pdf and joint pdf for the frequency of occurrence of rain events under various areal categories and Fig. 3b shows the corresponding change. In addition, the numbers in the square bracket show change in the mean yearly absolute frequency in Fig. 3b.
Left panel shows fractional frequency (colored contours, probability distribution function (pdf)) contributed by various areal categories (x-axis, refer Table 1) of daily CRA to seasonal and annual total. The numbers show joint probability distribution with respect to annual total. Right ordinate depicts the fraction of frequency of occurrence of seasonal CRA with respect to the observed annual total. Right panel shows change in the frequency of CRA (color contour) in recent past with respect to the base period. The numbers in square bracket show mean yearly absolute number of CRA or absolute change. Left ordinate (for both a and b) shows the season. Whereas right ordinate in right panel shows a fraction of seasonal change with respect to the base period. The asterisk (*) denotes a significant change (compared to inter-annual variability using 2-tailed t test) at 95% confidence level.
The average number of rain events during the base period is observed to be more than 7000/year. The monsoon season witnesses the maximum (49.7%) number of rain events followed by pre-monsoon (~24.8%), post-monsoon (12.8%) and winter (12.7%).
In the pre-monsoon, the areal category small (~isolated convective systems) and medium (~MCS) contributed collectively to more than 90% of the rain events. The leading rain-delivering season, monsoon, receives a maximum amount of rainfall (~70% of the annual) from the very large areal category (~synoptic systems), however, in terms of frequency it is only about ~1% of the annual rain events.
The maximum number of rain events in post-monsoon belongs to the areal category small (~70%) followed by medium (~20%), large (~7–10%), and very large (~0.5%). The corresponding contribution to the annual frequency in this season is 11.2%, 3.44%, 1.19%, and 0.08% respectively. The annual frequency distribution is specified in the top row of Fig. 3a. It shows that more than 70% of the annual rain events belong to areal category small and thus highlights its importance for rain-fed agricultural practices.
Figure 3b describes the change in the frequency of rain events in the recent past with respect to (w.r.t.) the base period. The numbers in the square bracket show the yearly mean actual number of rain events increased or decreased annually. Annually (top-most right ordinate), there is a ~3% decrease in the number of rain events although insignificant. This decrease is contributed from all the areal categories (top row) except the category large, which shows a significant ~10% increase w.r.t. the same category in the base period. Seasonally (right ordinate), a decrease in the number of rain events is observed in the winter (−13.1%),pre-monsoon (−3.7%), and monsoon (−0.9%). Whereas, post-monsoon shows a marginal insignificant increase (less than 1%) in the number of rain events..
Left panel show fractional volume (colored contours, probability distribution function (pdf)) contributed by various areal categories (x-axis, refer Table 1) of daily CRA to seasonal and annual total. The numbers show joint probability distribution with respect to annual total. Right ordinate depicts the fraction of seasonal volume contributed by CRA with respect to the observed annual total. Right panel show change in the frequency of CRA (color contour) in recent past with respect to the base period. Whereas right ordinate in right panel shows a fraction of seasonal change with respect to the base period. The asterisk (*) denotes a significant change (compared to inter-annual variability using 2-tailed t test) at 95% confidence level. Left ordinate (for both a and b) shows the season.
Seasonally, in the pre-monsoon season, both large (MCC) and very large (synoptic) systems have increased with a statistically significant increase of the former. During the monsoon season, the maximum rain-bearing category, very large shows a significant decrease followed by the category medium and small. The only exception is the category large which shows a significant increase in the number of rain events w.r.t. the base period. The post-monsoon season shows insignificant changes under all the categories although an appreciable decrease in the number, under the areal category very large.
Volume
Figure 4 shows the climatology of the base period and change observed in the recent past for spatially aggregated volume contribution. The fractional volume contribution (in percentage) by different areal categories of rain events to the seasonal and annual rainfall is shown in Fig. 4a
In the base period, the mean annual rainfall over the Indian land is about ~113 cm/year. Volume contribution to the annual rainfall by different seasons namely, winter, pre-monsoon, monsoon and post-monsoon are nearly 4%, 10.%, 75.% and 9% respectively (as shown in Fig. 4a right ordinate). Season-wise, the pre-monsoon season is the second-largest contributor to annual rainfall. The maximum amount of rainfall (more than 70%) in winter and pre-monsoon is contributed by the large size rain events. The highest rainfall contributor in the monsoon season is the synoptic events, which contributed more than 70% to the seasonal total and nearly 50% to the annual total. The event category large (MCC) in the monsoon season is the second-highest contributor (~21%) to the annual total. The post-monsoon season, which is also called the northeast monsoon season ranks third in the annual volume contribution. The areal category large contributes the maximum rainfall amount in the post-monsoon season and the corresponding annual contribution is 6%. The top row in Fig. 4a shows the annual volume distribution of rainfall contributed from different areal categories. Nearly 53% of the annual volume is contributed by very large events. Followed by large, medium and small areal categories (~39%, 6% and 1.5% respectively).
In the recent past (Fig. 4b), the winter and pre-monsoon seasons show an increase of 11.1% (non-significant) and 13.6% (significant) in rainfall amount respectively. Overall, the mean annual rainfall over Indian land is observed to have an insignificant marginal increase of 0.8%.
A significant increase is observed under the areal category large (MCC), resulting in an overall upturn of annual rainfall. Pre-monsoon is the dominant convection-driven rainfall season. In this season a significant increase of more than 10% under the areal category large is observed, followed by a ~40% increase (although insignificant) in the very large category (i.e., synoptic systems).
During the monsoon season, the topmost (70%) rain-delivering areal category, very large, shows a decrease of less than 5%. The second highest rain-delivering category, large, is observed to have a significant increase of more than 20% in volume contribution compared to the same areal category in the base period. The post-monsoon season shows an overall decline in rainfall with a majority of areal categories delivering less rainfall than the base period. Overall, in all seasons MCC show a significant increase in the contribution by volume and hence highlight the importance of the better prediction of convective systems in models.
Event-wise average CRA intensity
The annual and seasonal distribution of event average rainfall intensity (here onwards average intensity) for different areal categories for the base (blue) and recent past (red), is described in Fig. 5a–e. Event average intensity is the ratio of the total amount of rain to that of the number of pixels (or grids) contributing to a particular rain event. Figure 5f depicts the change in average intensity in the recent period relative to the base period, considering the mean statistics. Changes significant at a 95% confidence level are marked with an asterisk.
The box edges (a–e) show the 25th and 75th percentile value of event average intensity attain by daily CRA while horizontal bar within the box shows the median value of the distribution. The two horizontal lines outside the box (lower and upper whiskers) show the 10th and 90th percentile value of the distribution respectively. The inverted triangle is the maximum value of event average intensity of CRA in a given period. The subplot (f) shows the respective change in the mean statistics of various distributions. The changes significant at 95% confidence level are marked with black asterisks (*).
Figure 5a–e, for all seasons and annually, the interquartile range of the average intensity shows a monotonic increase with an increase in areal extent. Variability in average intensity decreases with an increase in event size. Thus, there is high uncertainty attached to mitigating the consequences that arise from small events as compared to larger events. In the recent past, unprecedented rain intensity is observed for small-sized events in all seasons except post-monsoon. The areal category small is observed to have a wide range of average intensity, from very low (~0.1 mm/day/grid) to very heavy (more than 500 mm/day/grid in JJAS) with a majority (~90%) having an average intensity below ~10 mm/day/grid in all seasons. In general, in the recent past, the average intensity of small events has decreased compared to the base period and is depicted in Fig. 5f, (first column (red color)) considering mean statistics.
The rain events under the areal category medium have relatively higher intensity compared to small. Among different seasons there are marginal differences in the average intensity distribution of medium size rain events. Comparing the medium-sized events in two periods, the average intensity distribution shifted marginally upward towards higher intensity in all seasons in the recent past. This increase in the average intensity of medium size events is statistically significant (Fig. 5f, second column) for the pre-monsoon season. Collectively, the annual distribution also shows this increase but is significant only at a 90% confidence level.
The areal category large has relatively less variability and higher event average intensity compared to medium and small. Overall, a marginal increase in the average intensity of large events is observed in the recent past in all seasons. The very large rain events show a relatively symmetric and less variable distribution around the median compared to other categories. The majority of the events have average intensity in the range of 8–20 mm/day/grid. A statistically significant increase in the event average intensity of very large (~synoptic) events is observed in the monsoon season (last column, Fig. 5f).
Here, it is worth noting that in monsoon, the frequency of synoptic events has decreased but their intensity has increased resulting in an overall decrease in volume contribution. These characteristics echos with previous studies that there is the strengthening of the thermodynamic driver (affecting intensity by the increase in moisture due to warming31, and a decrease in dynamic drivers32 (affecting the frequency of occurrence of events) of synoptic events.
EHR event frequency
Since extremely heavy rain events (EHR, defined under the data and methodology section) are having a more direct impact on human life and their livelihood, the characteristics of these events are studied in more detail.
Figure 6 shows the spatial distribution of the change in the frequency of occurrence of EHR events. Annually the small-sized events are observed to be increased over many parts of India, namely eastern India, the Indo-gangatic plain, south of western India and some parts near the east coast. The EHR with size medium is observed to have increased over some parts of north-east India and the south-east coast. A large number of locations in the eastern, northeast, and interiors of the southern peninsula shows an increase in the number of large-sized (~MCC) EHR events. Similarly many parts of central and eastern India show an increase in the frequency of very large (~synoptic) type EHR events that echo with the findings of the study by Roxy and coauthors15. Again to assert some more conclusions spatially aggregated analysis is presented in the following sections.
The black stippling denotes a significant change (compared to inter-annual variability using 2-tailed t-test) at 95% confidence level. For clarity, unit is represented in decadal change, however computed annually.
Figure 7 depicts spatially aggregated climatology and change in the frequency of occurrence of extreme rain events over India. In the base period (Fig. 7a), nearly 25 extreme rain events occurred every year. The seasonal distribution shows that more than 80% of these events occur during the monsoon season. A notable fraction of extreme events are also observed during pre-monsoon and post-monsoon seasons. According to the areal categories, more than half of the extreme rain events are associated with very large events (~synoptic systems) followed by ~38% of the events linked to the areal category large (MCC). The season winter contributes the least (~3%) to the annual extreme rain events. Here it is worth mentioning that, the areal category small also has a positive contribution towards extreme rain events, which shows its non-negligible importance in weather prediction and disaster management related to potential flash floods.
Left panel show fractional frequency (colored contours, probability distribution function (pdf)) contributed by various areal categories (x-axis, refer Table 1) of daily EHR CRA to seasonal and annual total. The numbers show joint probability distribution with respect to annual total. Right ordinate depicts the fraction of frequency of occurrence of seasonal EHR CRA with respect to the observed annual total. Right panel show change in the frequency of EHR CRA (color contour) in recent past with respect to the base period. The numbers in square bracket show mean yearly absolute number of EHR CRA or absolute change. Left ordinate (for both a and b) shows the season. Whereas right ordinate in right panel shows a fraction of seasonal change with respect to the base period. The asterisk (*) denotes a significant change (compared to inter-annual variability using 2-tailed t test) at 95% confidence level.
Figure 7b demonstrates the change in the frequency of extreme events in the recent past w.r.t. the base period. Annually, a significant ~55% increase in the occurrence of extreme events is observed. Seasonally, the extreme events in the pre-monsoon have nearly doubled followed by a significant increase of ~56% in monsoons. Post-monsoon and winter seasons are observed to have ~40% and 12.5% increases in the yearly frequency of extreme events respectively. These observations show that a detailed study of rain events occurring beyond the monsoon season is essential to improve the overall prediction of extreme rainfall events over India.
Aggregated annually, the categories large and very large are observed to have a significant increase in the frequency of occurrence of extreme rain events. Seasonally, a significant increase in the frequency of occurrence of areal category large (MCC) is observed for almost all seasons except in the winter season. During monsoons, a significant increase in the frequency of occurrence of the category very large is also observed.
EHR event average intensity
The spatially aggregated distribution of event average intensity for extreme events in various seasons and annually are shown in Fig. 8a–e. Subsection 8(f) shows the change in the mean statistics of event average intensity in the recent past w.r.t. the base period. In the base period for almost all seasons (except pre-monsoon), the areal category medium (Fig. 8a–e, blue box and whisker) is observed to have a maximum value of event average intensity followed by small, large and very large event types. Annually (Fig. 8e), the maximum value of event average intensity in the base period is observed to be contributed from the areal category medium having a value of more than 200 mm/day.
The box edges show 25th and 75th percentile value of event average intensity attain by EHR while horizontal bar within the box shows the median value of the distribution. The two horizontal lines outside the box (lower and upper whiskers) show 10th and 90th percentile value of the distribution respectively. The inverted triangle is the maximum value of event average intensity of EHR in a given period. The subplot (f) shows the respective change in the mean statistics of various distributions. The changes significant at 95% confidence level are marked with black asterisks (*).
Comparing the recent past to the base period (Fig. 8f), the areal category small is observed to be significantly intensified by more than 50% in all seasons with maximum intensification (~200%) during monsoon. The event category medium is observed to be weakening significantly in all seasons. The collective distribution shows an annual intensification of small event types, which is observed to be contributed significantly from all seasons. The areal type very large is noted to have the least variability in average intensity relative to other categories. Overall, in the recent past, unprecedented values of the event average intensity are recorded for small EHR with respect to the base period in all seasons. Thus, this observation highlights the need for process studies beyond synoptic scale events contributing from all seasons.
Discussion
The climatology (1951–1980) and changes in recent period (1981–2020) of the CRA over India are studied, including the information related to interannual variability. The scope of the present study does not considered tracking of the CRA in space and time but, this could be of importance to study the dynamics of the system. The seasonal trend analysis very well echoes with a majority of the past studies for the monsoon season but at the same time, the observed reduction in the magnitude of the annual rainfall trend highlights the compensating effect of the rainfall contribution from other seasons to lessen the severity of widespread monsoon season drying. Thus, if characteristics from all seasons are taken into consideration, water management in the region may get better.
The comparison of the annual mean statistics of the two periods shows an insignificant change in the frequency of occurrence and volume. The seasonal volume contribution to annual rainfall is mainly contributed from the monsoon season (~75%) followed by pre-monsoon (~10%), post-monsoon (~9%) and winter (~4%). However, the number of daily rain events in each season contributing to the above-mentioned amount has much less disparity (i.e., ~50%, 25%, 13% and 13% for monsoon, pre-monsoon, post-monsoon and winter respectively). Thus, to support the agricultural practices throughout the year in regions with no or limited facility for artificial irrigation, understanding and hence improvement in the predictability of daily rain events for all seasons is likely to play a vital role.
Despite the low frequency of occurrence (~1% of annual rain events) of very large events (i.e., synoptic), it contributes a maximum amount of rainfall to the annual total (~53%) and is associated with a stronger event average intensity. A significant decrease in the frequency of these very large events is observed during monsoon which echoes with the previous findings. In general, the frequency (Fig. 3b) and volume (Fig. 4b) of rain delivered by event category large (~MCC) have increased significantly in almost all seasons that also echoes with findings of some recent studies33,34, and hence, highlights the importance of understanding and improvement required in the convective parameterization, in the models for better future projections of rainfall.
Further, it is noted that more than 70% of the rain events in all seasons are small in size and contribute to nearly ~1% of annual rainfall. Even though the volume contribution from the small event is only 1% of the annual total, its frequent occurrence and large variability in event average intensity make it an important predictive component to reduce the uncertainties in disaster management and rain-fed agricultural practices.
Considering, the significant socio-economic impact of EHR events, it is found that the majority of highest intensity events (EHR, Fig. 8e) are small scale which echoes with the finding of a past study35. In addition, the increase in the frequency and significant intensification (~50–200%) of EHR small events in all seasons demand a better understanding of rain-bearing processes beyond the synoptic-scale processes of monsoon season. The other EHR finding is that CRA of size of the order of MCC has increased significantly in all seasons which correspondence to the findings of the past studies15,25, that large collocated extreme events have increased since 1980. However, here it is important to note that these past studies studied only monsoon season and focused largely on central India whereas the present study considered all of India, annual rainfall distribution. The other tangible feature of the climatological aspect of this study would provide threshold values for the object-orientated rainfall verification method. This study utilizes the concept of contiguous rain area (CRA), which was initially used for object-oriented rainfall verification36. In the verification method, threshold values of rain intensity and a minimum number of connected rain grid points play a significant role. However, the successor studies utilizing this verification method for the Indian region chose the threshold values fairly subjectively37,38. Thus, to have better empirical estimates of threshold values for precipitation verification purposes, the present study may add value.
In conclusion, for better water management and hazard mitigation under the changing climate and land-use scenario, process studies of rain-bearing systems of sizes smaller than synoptic scale and seasons beyond monsoon season need urgent attention.
Methods
The widely used, Indian meteorological department’s daily gridded rain gauge dataset39 at quarter-degree resolution for the period 1951 to 2020 is used for this study. The data set is based on daily rainfall records from 6955 rain gauge stations spread across India from 1901 to 2020 and is the highest number of stations used by any gridded data sets currently available over India. The data density varies from about 1500 from the beginning to about 4000 by the end of the century. For the current analysis, the data for the period 1950–2020 is used for which the data density was relatively higher, that is more than 3100 stations per day. It may be noted that time series analysis using a changing rain-gauge network may induce artificial trends, compared to a fixed network of rain gauges. However, part of our results are comparable with many earlier studies (including a study with fixed gauge data20 and thus provide confidence in the rest of the results, obtained from the present study.
All the data is analyzed for annual (Ann) and four different seasons namely winter (December, January, February (DJF)), pre-monsoon (March, April, May (MAM), monsoon (June, July, August, September (JJAS)), and post-monsoon (October and November (ON).
Study period
In the context of South Asia including India, the implementation of big irrigation schemes and hydro-power dams evolved during the 1950s and1960s40. The land-use change in India, in the recent past, has primarily occurred because of urbanization41 and intensive agriculture/irrigation practices42,43. The literature suggests that in India, urbanization21,44,45 and artificial irrigation practices owing to the green revolution40,46 have intensified during the decade of 1971–1981 and have a substantial role in modulating the rainfall41,47,48. Also, the aerosol content in the atmosphere over India49 has a noticeable increase post-1970. Considering these facts, the study period is divided into two parts: the base period of 30 years from 1951 to 1980 is the period when the majority of the infrastructure was built (i.e., the land-use change) and the recent past consists of 40 years from 1981 to 2020, which depicts the impact of pre-built infrastructure in modulating rainfall characteristics and vice versa and thus the susceptibility of rainfall to land use and climate change.
CRA method
Daily rain events are identified as CRA36 objects with non-zero values. For holistic and more inclusive interpretation, a full spectrum of rain event sizes, without any threshold value for rain rate (all seasons except monsoon for which a threshold of 1 mm is considered to reduce error in continuity introduce by interpolation) is considered. Then the number of contiguous non-zero value pixels is identified as a rain event which is usually done in the object-oriented precipitation verification method. Thus the above method resulted in the smallest event size being one isolated pixel with a rainfall rate greater than zero (except for monsoon season).
In literature, the definition of extreme events is majorly divided into two groups: (1) based on percentile values and (2) based on absolute values. Both have their merits and demerits. However, according to WMO guidelines50, fixed absolute values represent the disaster-causing potential (floods, in the current context) better than percentile values. Thus, in this study, we have described the characteristics and the associated change in daily Extremely Heavy Rain events (EHR) in absolute numbers. An event is called EHR, if a rain event has at least one grid point with a value greater than or equal to 244.5 mm/day. This threshold is based on the classification defined by India Meteorological Department (SOP, 2010)51.
Data availability
IMD daily gridded rainfall data (mm/day) at quarter degree resolution can be obtained from India Meteorological Department’s Pune website https://www.imdpune.gov.in/lrfindex.php for the period 1951–2020.
Code availability
The analysis codes are available from the given repository “Recent Changes in the Climatological Characteristics of Daily Contiguous Rain Areas over India Mansi Bhowmick, Sandeep Sahany, Ananda Kumar Das”, Mendeley Data, V1, doi: 10.17632/4vj3zyy5d6.1.
References
Agricultural statistics: agricultural statistics at a glance, ministry of agriculture & farmers welfare, Government of India. https://agricoop.gov.in/Documents/agristatglance2018.pdf (2018).
Amrith, S. S. Risk and the South Asian monsoon. Clim. Change 151, 17–28 (2018).
Gadgil, S. & Gadgil, S. The Indian Monsoon, GDP and agriculture. Econ. Polit. Wkly. 41(Nov), 4887–4895 (2006).
Krishna Kumar, K., Rupa Kumar, K., Ashrit, R. G., Deshpade, N. R. & Hansen, J. W. Climate impact on India agriculture. Int. J. Climatol. 24, 1375–1393 (2004).
Rajeevan, M., Unnikrishnan, C. K., Kumar, K. N. & Sreekala, P. P. Northeast monsoon over India: variability and prediction. Meteorol. Appl. 19, 226–236 (2012).
Jayaraman, T. Climate change and agriculture: a review article with special reference to India. Rev. Agrar. Stud. 2, 17–78 (2011).
Romatschke, U. & Houze, R. A. Jr. Characteristics of precipitating convective systems in the premonsoon season of South Asia. J. Hydrometeorol. 12, 157–180 (2011).
Das, S., and Coauthors: The SAARC STORM: a coordinated field experiment on severe thunderstorm observations and regional modeling over the South Asian region. Bull. Am. Meteorol. Soc. 95, 603–617 (2014).
Sreekala, P. P., Vijaya, S., Rao, B. & Rajeevan, M. Northeast monsoon rainfall variability over south peninsular India and its teleconnections. Theor. Appl. Climatol. 108, 73–83 (2012).
Tyagi, A. Thunderstorm climatology over Indian region. Mausam 58, 189–212 (2007).
RoyBhowmik, S. K., SenRoy, S. & Kundu, P. K. Analysis of large-scale conditions associated with convection over the Indian monsoon region. Int. J. Climatol. 28, 797–821 (2008).
Guhathakurta, P. & Rajeevan, M. Trends in the rainfall pattern over India. Int. J. Climatol. 28, 1453–1469 (2008).
Dash, S. K., Kulkarni, M. A., Mohanty, U. C. & Prasad, K. Changes in the characteristics of rain events. India J. Geophys. Res. 114, D10109 (2009).
Ghosh, S., Luniya, V. & Gupta, A. Trend analysis of Indian summer monsoon rainfall at different spatial scales. Atmos. Sci. Lett. 10, 285–290 (2009).
Roxy, M. K. et al. A threefold rise in widespread extreme rain events over central India. Nat. Commun. 8, 708 (2017).
Nikumbh, A. C., Chakraborty, A., Bhat, G. S. & Frierson, D. M. W. Large-scale extreme rainfall-producing synoptic systems of the Indian summer monsoon. Geophys. Res. Lett. 47, e2020GL088403 (2020).
Turner, A. G. & Annamalai, H. Climate change and the South Asian summer monsoon. Nat. Clim. Change 2, 587–595 (2012).
Mohanty, U. C. et al. Indian monsoon variability and extreme weather events: recent improvements in observations and modelling. Proc. Indian Natl Sci. Acad. 86, 503–524 (2019).
Goswami, B. N., Venugopal, V., Sengupta, D., Madhusoodanan, M. S. & Xavier, P. K. Increasing trend of extreme rain events over India in a warming environment. Science 314, 1442–1445 (2006).
Rajeevan, M., Bhate, J. & Jaswal, A. K. Analysis of variability and trends of extreme rainfall events over India using 104 years of gridded daily rainfall data. Geophys. Res. Lett. 35, L18707 (2008).
Vittal, H., Karmakar, S. & Ghosh, S. Diametric changes in trends and patterns of extreme rainfall over India from pre‐1950 to post‐1950. Geophys. Res. Lett. 40, 3253–3258 (2013).
Sen Roy, S. & Balling, R. C. Trends in extreme daily precipitation indices in India. Int. J. Climatol. 24, 457–466 (2004).
Guhathakurta, P., Sreejith, O. P. & Menon, P. A. Impact of climate change on extreme rainfall events and flood risk in India. J. Earth Syst. Sci. 120, 359–373 (2011).
Karunasagar S., Rajeevan M. and Rao. S.: On increasing monsoon rainstorms over India. Nat. Hazards. 84 https://doi.org/10.1007/s11069-016-2662-9 (2017).
Nikumbh, A. C., Chakraborty, A. & Bhat, G. S. Recent spatial aggregation tendency of rainfall extremes over India. Sci. Rep. 9, 10321 (2019).
Dhar, O. N. & Nandargi, S. On some characteristics of severe rainstorms of India. Theor. Appl. Climatol. 50, 205–212 (1995).
Orlanski, I. A rationale subdivision of scales for atmospheric processes. Bull. Am. Meteorol. Soc. 56, 527–530 (1975).
Mann, H. B. Nonparametric tests against trend. Econometrica: J. Econom. Soc. 13, 245–259 (1945).
Koteswarm, P. & Srinivasan, V. Thunderstorm over Gangetic West Bengal in the pre-monsoon season and the synoptic factors favourable for their formation. Indian J. Met. Geophys. 9, 301–312 (1958).
Tyagi, A., Sikka, D. R., Goyal, S. & Bhowmick, M. A satellite based study of pre-monsoon thunderstorms (Nor’westers) Over Eastern India and their organization into mesoscale convective complexes. Mausam 63, 29–54 (2012).
Sørland, S. L. & Sorteberg, A. Low-pressure systems and extreme precipitation in central India: sensitivity to temperature changes. Clim. Dyn. 47, 465–480 (2016).
Vishnu, S., Francis, P. A., Shenoi, S. S. C. & Ramakrishna, S. S. V. S. On the decreasing trend of the number of monsoon depressions in the Bay of Bengal. Environ. Res. Lett. 11, 014011 (2016).
Sreenath, A. V., Abhilash, S., Vijayakumar, K. & Mapes, B. West coast India’s rainfall is becoming more convective. npj Clim. Atmos. Sci. 5, 36 (2022).
Sreenath A. V., Abhilash S., Ajilesh P.P.: Changes in the dynamical, thermodynamical and hydrometeor characteristics prior to extreme rainfall events along the southwest coast of India in recent decades. Atm. Res. https://doi.org/10.1016/j.atmosres.2023.106752 (2023).
Moron, V., Barbero, R., Fowler, H. J. & Mishra, V. Storm types in India: linking rainfall duration, spatial extent and intensity. Philos. Trans. R. Soc. A. 379, 20200137 (2021).
Ebert, E. E. & McBride, J. L. Verification of precipitation in weather systems: determination of systematic errors. J. Hydrol. 239, 179–202 (2000).
Das, A. K., Bhowmick, M., Kundu, P. K. & Roy Bhowmik, S. K. Verification of WRF rainfall forecasts over India during monsoon 2010: CRA method. Geofizika 31(2), 105–126 (2014).
Sharma, K. et al. Assessment of Met office Unified Model (UM) quantitative precipitation forecasts during the Indian summer monsoon: Contiguous Rain Area (CRA) approach. J. Earth Syst. Sci. 128, 4 (2019).
Pai, D. S. et al. Development of a new high spatial resoultion (0.25 × 0.25) long period (1901–2010) daily gridded rainfall data set over India and its comparison with existing data sets over the region. Mausam 65, 1–18 (2014).
Reddy, V. R., Saharawat, Y. S. & George, B. Watershed management in South Asia: a synoptic review. J. Hydrol. 551, 4–13 (2017).
Kishtawal, C. M., Niyogi, D., Tewari, M., Pielke, R. A. Sr. & Shepherd, J. M. Urbanization signature in the observed heavy rainfall climatology over India. Int. J. Climatol. 30, 1908–1916 (2010).
Douglas, E. M. et al. Changes in moisture and energy fluxes due to agricultural land use and irrigation in the Indian Monsoon belt. Geophys. Res. Lett. 33, L14403 (2006).
Douglas, E. M., Beltra´n-Przekurat, A., Niyogi, D., Pielke, R. A. Sr & Vorosmarty, C. J. The impact of agricultural intensification and irrigation on land-atmosphere interactions and Indian monsoon precipitation: A mesoscale modeling perspective. Glob. Planet. Change 67, 117–128 (2009).
Crook, N. & Dyson, T. Urbanization in India: results of the 1981 census. Popul. Dev. Rev. 8, 145–155 (1982).
Mohan, R. & Pant, C. Morphology of urbanization in India: Some results from 1981 census. Econ. Pol. Wkly 17, 1579–1588 (1982).
Roy, S. S. et al. Impacts of the agricultural green revolution-induced land use changes on air temperatures in India. J. Geophys. Res. 112, D21108 (2007).
Niyogi, D., Kishtawal, C., Tripathi, S. & Govindaraju, R. S. Observational evidence that agricultural intensification and land use change may be reducing the Indian summer monsoon rainfall. J. Water Resour. Res. 46, W03533 (2010).
Paul, S., Ghosh, S., Mathew, M., Devanand, A. & Karmakar, S. Increased spatial variability and intensification of extreme monsoon rainfall due to urbanization. Sci. Rep. 8, 3918 (2019).
Ramanathan, V. et al. Atmospheric brown clouds: Impacts on South Asian climate and hydrological cycle. Proc. Natl. Acad. Sci. USA 102, 5326–5333 (2005).
Tank A. K. M. C., Zwiers F. W. & Zhang X. Guidelines on analysis of extremes in a changing climate in support of informed decisions for adaptation. World Meteorol. Organ. WCDMP No. 72, pp 14–17 (2009).
Standard operation procedure-weather forecasting and warning published by india meteorological department, Ministry of Earth Sciences, Government of India, Mausam Bhawan, Lodi Road, New Delhi-110003 (2010).
The NCAR Command Language (Version 6.6.2) [Software]: Boulder, Colorado: UCAR/NCAR/CISL/TDD. https://doi.org/10.5065/D6WD3XH5 (2019).
Schumacher, R. S. & Rasmussen, K. L. The formation, character and changing nature of mesoscale convective systems. Nat. Rev. Earth Environ. 1, 300–314 (2020).
Laing, A. G. & Fritsch, J. M. Mesoscale convective complexes over the Indian monsoon region. J. Clim. 6, 911–919 (1993).
Maddox, R. A. Mesoscale convective complexes. Bull. Am. Meteor. Soc. 61, 1374–1387 (1980).
Moron, V., Robertson, A. W. & Pai, D. S. On the spatial coherence of sub-seasonal to seasonal Indian rainfall anomalies. Clim. Dyn. 49, 3403–3423 (2017).
Mooley, D. A. Some aspects of Indian monsoon depressions and the associated rainfall. Mon. Weather Rev. 101, 271–280 (1973).
Goswami, B. N., AjayMohan, R. S., Xavier, P. R. & Sengupta, D. Clustering of synoptic activity Indian summer monsoon intraseasonal oscillations. Geophys. Res. Lett. 30, 1431 (2003).
Mohapatra, M. & Mohanty, U. C. Some characteristics of low pressure systems and summer monsoon rainfall over Orissa. Curr. Sci. 87, 1245–1255 (2004).
Acknowledgements
The use of IMD gridded rainfall data and NCAR command language (NCL, version 6.6.2)52 software for figure generation is sincerely acknowledged. This study is partly supported by The DST Center of Excellence in Climate Modeling, CAS, IIT Delhi (project number: RP03350), India and the School of Earth and Environment, University of Leeds, UK. We sincerely thank Dr. J.L. McBride for the insightful discussion of the study. M.B. sincerely thank Mr. Kaushik Patel for providing technical support in developing the CRA code used in this study.
Author information
Authors and Affiliations
Contributions
M.B. has designed the study and contributed towards formal analysis, interpretation of the results and wrote the first draft. S.S. contributed towards conceptualization and the analysis design. A.K.D. contributed towards the writing and review of the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Bhowmick, M., Sahany, S. & Das, A.K. Recent changes in the climatological characteristics of daily contiguous rain areas over India. npj Clim Atmos Sci 6, 178 (2023). https://doi.org/10.1038/s41612-023-00464-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41612-023-00464-6
