Table 1 summarizes the major conclusions of past trend analysis studies. Broadly, studies had used the data for about 50–100 years, for the whole of India or a part thereof (river basins or states). In most cases, a simple trend analysis was carried out. As can be seen, the results are somewhat different and not always consistent. Inconsistency in the outcomes of the various studies could arise due to several reasons. Different studies may have used different data or the data that may have been compiled differently. The study area and its geographical coverage might have been different. Furthermore, due to changes in climate, land use/cover, and other causes, the statistical properties of meteorological data series may undergo changes. Such changes can be detected by statistical tests. In case the series has some statistically significant change points, trend analysis by using the whole series is not very useful as the statistical properties of the series before and after the change point will be different. Results of the statistical analysis presented in subsequent sections show the presence of change points which are significant at a 5% significance level. Some studies listed in Table 1 may not have tested the data series for the presence of change points and the results of such studies may not be consistent with others.
The key outcomes of some trend analysis studies
| Authors . | Study area . | Data period . | Outcome . |
|---|---|---|---|
| Singh et al. (2005) | Whole India | Basin-scale area-averaged rainfall (RF) series for 1871–2000 | Annual rainfall over Central Indian basins (Mahi, Sabarmati, Tapi, Narmada, Mahanadi and Godavari) had decreasing trend since the 1960s; some basins have an increasing trend: Ganga from 1993, Indus from 1954, Krishna from 1953, Cauvery from 1929, and Brahmaputra from 1998 |
| Rajeevan et al. (2006) | Whole India | Monthly, seasonal and annual RF of 36 meteorological sub-divisions, 1901–2003 | Break (active) periods during the monsoon season were identified and were comparable with those identified by earlier studies. No evidence was found for any statistically significant trends in the number of break or active days during the period 1951–2003 |
| Ramesh & Goswami (2007) | Whole India | 1951–2003 | Falling trends in early and late monsoon RF and the number of rainy days |
| Dash et al. (2007) | Whole India | Monthly rainfalls, 1871–2002 | Small increase in RF in winters (Jan and Feb), pre-monsoon (March to May) and post-monsoon (Oct. to Dec.). Summer monsoon RF showed small decreasing trend |
| Guhathakurta & Rajeevan (2008) | Whole India | Monthly, seasonal and annual RF of 36 meteorological sub-divisions, 1901–2003 | For summer monsoon season, RF in three sub-divisions had significant decreasing trend. Trends over eight sub-divisions were significant and rising. Contribution of June, July and Sept. RF to annual RF was decreasing for few sub-divisions while that of August RF was increasing in few other sub-divisions |
| Singh et al. (2008) | Nine basins in North-West and central India | Data from 43 stations, 90 to 100 years in 20th century. | Increasing trends in annual RF in majority of river basins, ranging from 2 to 19% of the mean per 100 years in over eight basins |
| Kumar & Jain (2011) | Whole India | Daily-gridded RF at 1° × 1° scale, for 1951–2004 | Annual RF: 15 of 22 basins showed a falling trend; only one significant. Six basins showed increasing trend, one significant. Monsoon RF increased over six basins; decreased over 16 basins, two significant. Annual rainy days: four basins had increasing (non-significant) trend; 15 basins had decreasing trend, three significant |
| Kumar et al. (2010) | Whole India | Monthly RF 1871–2005 (135 years) for 30 met sub-divisions. | Annual RF: 15 sub-div. showed rising trend, three significant; 15 sub-divisions showed falling trend, only one significant. Most sub-div. showed very little change in RF in non-monsoon months. Five main regions showed no significant trend in annual, seasonal and monthly RF. For whole India, no significant trend in annual, seasonal, or monthly RF |
| Pal & Al-Tabbaa (2011) | Whole India and Kerala state of India | 1954–2003 | Trends in seasonal precipitations have large regional variations. No significant trends found in annual or seasonal precipitation amount in various regions in India. Precipitation has an increasing tendency in winter and autumn seasons, decreasing tendency in spring and monsoon seasons in Kerala state |
| Subash et al. (2010) | Whole India | 1889–2008 | Annual, seasonal and monthly RF in five meteorological sub-divisions of Central North-East India showed a significant falling trend |
| Kumar & Jain (2011) | 22 river basins of India | 1951–2004 | Fifteen river basins had decreasing trend in annual rainfall; one basin had significant decreasing trend at 95% confidence level. Six basins had increasing rainfall trend, one basin showed a significant positive trend. Monsoon rainfall increased over six basins, decreased over 16 basins and decreasing trend for two basins was statistically significant. Four river basins experienced increasing (non-significant) trend in annual rainy days; 15 basins had a decreasing trend in annual rainy days; such trend in three basins was statistically significant |
| Subash & Sikka (2014) | Data of 36 meteorological sub-divisions of India | 1904–2003 | Increasing trend was seen in annual RF in all the homogeneous regions, except NE |
| Taxak et al. (2014) | Wainganga river basin in India | 0.5° × 0.5° resolution gridded rainfall for the period 1901–2012. | Most grids showed falling annual RF, decreasing trend was significant only in seven grids. Increasing trend was observed in post-monsoon season but was not significant. Seven grids showed significant falling trend in monsoon RF. Reported 8.45% fall in annual RF during 1901–2012. 1948 was the most probable change year. Rising trend in RF in Wainganga basin during 1901–1948 but falling trend during 1949–2012 |
| Pingale et al. (2014) | 33 urban centers in Rajasthan state of, India | 1971–2005 | Mann–Kendall test and Sen's slope estimator were used to examine trends in urban centers in an arid area. Both rising and falling trends were found. No geographical trend could be detected |
| Panda & Sahu (2019) | Kalahandi, Bolangir and Koraput districts, Odisha | 1980–2017 | Examined long-term changes and short-term fluctuations in monsoon RF by using MK test and Sen's slope estimator. Statistically significant trends were detected. RF data showed a quite good increasing trend (Sen's slope = 4.034) for monsoon season |
| Praveen et al. (2020) | Thirty-four meteorological sub-divisions | Annual and seasonal rainfall data of 1901–2015 (115 years) | Annual and seasonal variability was highest in sub-divisions (SD) of Western India, the lowest in Eastern and North India. MK test results showed that SDs of NE, South and Eastern India had significant negative trend, while SDs like Sub-Himalayan Bengal, Gangetic Bengal, Jammu & Kashmir, Konkan & Goa, and Marathwada showed positive trend. Change points were detected in between 1950–1966. Most SDs had increased variability and significant negative trend after change point |
| Saini et al. (2020) | West Coast Plain and Hill Agro-Climatic Region of Western Ghats region | 1901–2017 (117 years). | Employed Modified Mann–Kendall's test, Linear Regression, Innovative Trend Analysis, Sen's Slope test, Weibull's Recurrence Interval, and other statistical techniques. RF trend was significant and falling for the months Jan., July, and Winter season; Aug., Sept. and winter season showed an increasing trend |
| Singh et al. (2021) | 36 districts of Maharashtra State of India, | 1901–2018 (118 years) | Significantly falling trends for pre-monsoon and winter RF in many districts. Both annual and seasonal RFs had rising and falling trends. Out of 185 series analyzed, 168 had some trend |
| Bora et al. (2022) | Seven states of NE India | 1901–2020 | Modified MK test results revealed that annual RF in Assam and Nagaland showed negative trends at a 99% significance level, Meghalaya and Mizoram showed a positive trend at a 99% significance level, while Arunachal Pradesh, Manipur and Tripura showed no significant trends |
| Extreme rainfall | |||
| Joshi & Rajeevan (2006) | India | Rainfall data for 1901–2000, for 100 stations | Most of the extreme rainfall indices for annual period and for southwest monsoon showed significant rising trends over Northwestern parts of Peninsula and the West Coast |
| Rajeevan et al. (2008) | India | High resolution daily gridded data for 1901–2004, 104 years | Statistically significant long-term trend amounting to 6% per decade in frequency of extreme RF events |
| Pattanaik & Rajeevan (2010) | India | Long-term trend in extreme monsoon rainfall for 1951–2005 | Average frequency of extreme RFs along with contribution of such events to monsoon season has shown a significant increasing trend. Rising contribution from extreme RF events is countered by falling trend in low RF events |
| Authors . | Study area . | Data period . | Outcome . |
|---|---|---|---|
| Singh et al. (2005) | Whole India | Basin-scale area-averaged rainfall (RF) series for 1871–2000 | Annual rainfall over Central Indian basins (Mahi, Sabarmati, Tapi, Narmada, Mahanadi and Godavari) had decreasing trend since the 1960s; some basins have an increasing trend: Ganga from 1993, Indus from 1954, Krishna from 1953, Cauvery from 1929, and Brahmaputra from 1998 |
| Rajeevan et al. (2006) | Whole India | Monthly, seasonal and annual RF of 36 meteorological sub-divisions, 1901–2003 | Break (active) periods during the monsoon season were identified and were comparable with those identified by earlier studies. No evidence was found for any statistically significant trends in the number of break or active days during the period 1951–2003 |
| Ramesh & Goswami (2007) | Whole India | 1951–2003 | Falling trends in early and late monsoon RF and the number of rainy days |
| Dash et al. (2007) | Whole India | Monthly rainfalls, 1871–2002 | Small increase in RF in winters (Jan and Feb), pre-monsoon (March to May) and post-monsoon (Oct. to Dec.). Summer monsoon RF showed small decreasing trend |
| Guhathakurta & Rajeevan (2008) | Whole India | Monthly, seasonal and annual RF of 36 meteorological sub-divisions, 1901–2003 | For summer monsoon season, RF in three sub-divisions had significant decreasing trend. Trends over eight sub-divisions were significant and rising. Contribution of June, July and Sept. RF to annual RF was decreasing for few sub-divisions while that of August RF was increasing in few other sub-divisions |
| Singh et al. (2008) | Nine basins in North-West and central India | Data from 43 stations, 90 to 100 years in 20th century. | Increasing trends in annual RF in majority of river basins, ranging from 2 to 19% of the mean per 100 years in over eight basins |
| Kumar & Jain (2011) | Whole India | Daily-gridded RF at 1° × 1° scale, for 1951–2004 | Annual RF: 15 of 22 basins showed a falling trend; only one significant. Six basins showed increasing trend, one significant. Monsoon RF increased over six basins; decreased over 16 basins, two significant. Annual rainy days: four basins had increasing (non-significant) trend; 15 basins had decreasing trend, three significant |
| Kumar et al. (2010) | Whole India | Monthly RF 1871–2005 (135 years) for 30 met sub-divisions. | Annual RF: 15 sub-div. showed rising trend, three significant; 15 sub-divisions showed falling trend, only one significant. Most sub-div. showed very little change in RF in non-monsoon months. Five main regions showed no significant trend in annual, seasonal and monthly RF. For whole India, no significant trend in annual, seasonal, or monthly RF |
| Pal & Al-Tabbaa (2011) | Whole India and Kerala state of India | 1954–2003 | Trends in seasonal precipitations have large regional variations. No significant trends found in annual or seasonal precipitation amount in various regions in India. Precipitation has an increasing tendency in winter and autumn seasons, decreasing tendency in spring and monsoon seasons in Kerala state |
| Subash et al. (2010) | Whole India | 1889–2008 | Annual, seasonal and monthly RF in five meteorological sub-divisions of Central North-East India showed a significant falling trend |
| Kumar & Jain (2011) | 22 river basins of India | 1951–2004 | Fifteen river basins had decreasing trend in annual rainfall; one basin had significant decreasing trend at 95% confidence level. Six basins had increasing rainfall trend, one basin showed a significant positive trend. Monsoon rainfall increased over six basins, decreased over 16 basins and decreasing trend for two basins was statistically significant. Four river basins experienced increasing (non-significant) trend in annual rainy days; 15 basins had a decreasing trend in annual rainy days; such trend in three basins was statistically significant |
| Subash & Sikka (2014) | Data of 36 meteorological sub-divisions of India | 1904–2003 | Increasing trend was seen in annual RF in all the homogeneous regions, except NE |
| Taxak et al. (2014) | Wainganga river basin in India | 0.5° × 0.5° resolution gridded rainfall for the period 1901–2012. | Most grids showed falling annual RF, decreasing trend was significant only in seven grids. Increasing trend was observed in post-monsoon season but was not significant. Seven grids showed significant falling trend in monsoon RF. Reported 8.45% fall in annual RF during 1901–2012. 1948 was the most probable change year. Rising trend in RF in Wainganga basin during 1901–1948 but falling trend during 1949–2012 |
| Pingale et al. (2014) | 33 urban centers in Rajasthan state of, India | 1971–2005 | Mann–Kendall test and Sen's slope estimator were used to examine trends in urban centers in an arid area. Both rising and falling trends were found. No geographical trend could be detected |
| Panda & Sahu (2019) | Kalahandi, Bolangir and Koraput districts, Odisha | 1980–2017 | Examined long-term changes and short-term fluctuations in monsoon RF by using MK test and Sen's slope estimator. Statistically significant trends were detected. RF data showed a quite good increasing trend (Sen's slope = 4.034) for monsoon season |
| Praveen et al. (2020) | Thirty-four meteorological sub-divisions | Annual and seasonal rainfall data of 1901–2015 (115 years) | Annual and seasonal variability was highest in sub-divisions (SD) of Western India, the lowest in Eastern and North India. MK test results showed that SDs of NE, South and Eastern India had significant negative trend, while SDs like Sub-Himalayan Bengal, Gangetic Bengal, Jammu & Kashmir, Konkan & Goa, and Marathwada showed positive trend. Change points were detected in between 1950–1966. Most SDs had increased variability and significant negative trend after change point |
| Saini et al. (2020) | West Coast Plain and Hill Agro-Climatic Region of Western Ghats region | 1901–2017 (117 years). | Employed Modified Mann–Kendall's test, Linear Regression, Innovative Trend Analysis, Sen's Slope test, Weibull's Recurrence Interval, and other statistical techniques. RF trend was significant and falling for the months Jan., July, and Winter season; Aug., Sept. and winter season showed an increasing trend |
| Singh et al. (2021) | 36 districts of Maharashtra State of India, | 1901–2018 (118 years) | Significantly falling trends for pre-monsoon and winter RF in many districts. Both annual and seasonal RFs had rising and falling trends. Out of 185 series analyzed, 168 had some trend |
| Bora et al. (2022) | Seven states of NE India | 1901–2020 | Modified MK test results revealed that annual RF in Assam and Nagaland showed negative trends at a 99% significance level, Meghalaya and Mizoram showed a positive trend at a 99% significance level, while Arunachal Pradesh, Manipur and Tripura showed no significant trends |
| Extreme rainfall | |||
| Joshi & Rajeevan (2006) | India | Rainfall data for 1901–2000, for 100 stations | Most of the extreme rainfall indices for annual period and for southwest monsoon showed significant rising trends over Northwestern parts of Peninsula and the West Coast |
| Rajeevan et al. (2008) | India | High resolution daily gridded data for 1901–2004, 104 years | Statistically significant long-term trend amounting to 6% per decade in frequency of extreme RF events |
| Pattanaik & Rajeevan (2010) | India | Long-term trend in extreme monsoon rainfall for 1951–2005 | Average frequency of extreme RFs along with contribution of such events to monsoon season has shown a significant increasing trend. Rising contribution from extreme RF events is countered by falling trend in low RF events |