Trends in Precipitation Extremes over India - (IMD), Pune

OCTOBER2006NCC3NCC RESEARCH REPORTTrends in PrecipitationExtremes over IndiaU. R. JoshiM. RajeevanNATIONAL CLIMATE CENTREOFFICE OF THEADDITIONAL DIRECTOR GENERAL OF METEOROLOGY (RESEARCH)INDIA METEOROLOGICAL DEPARTMENTPUNE - 411 005INDIAMETEOROLOGICALNAT IONALCLIMATEDEPARTMENTCENT RE

National Climate CentreResearch Report No: 3/2006Trends in Precipitation Extremes over IndiaU. R. Joshi and M. RajeevanNational Climate CentreIndia Meteorological DepartmentPUNE. INDIA411005ncc@imdpune.gov.in

AbstractOne of the most significant consequences of global warming due to increase ingreenhouse gases would be an increase in magnitude and frequency of extremeprecipitation events. In the present study the trends in extreme rainfall indices for theperiod 1901-2000 were examined for 100 stations over India. The trends for thesouthwest monsoon season and annual period were calculated separately. Theresults show that most of the extreme rainfall indices have shown significant positivetrends over the west coast and northwestern parts of Peninsula. However, two hillystations considered (Shimla and Mahabaleshwar) have shown decreasing trend insome of the extreme rainfall indices.1

1. IntroductionAbout 60-90% of the annual rainfall over India is received during thesouthwest monsoon season (June to September), which is vital for the economy ofthe country. Inter-annual variation of seasonal and annual rainfall is a subject formore serious research work in India. However, information about the long termtrends of rainfall is also important. Previous studies have addressed the issue ofchanges in the mean rainfall. For example, Guhathakurta and Rajeevan (2006) haveshown that there is no long term trend in the southwest monsoon seasonal rainfallover the country as a whole, but there are significant regional variations. However,changes in extreme precipitation are also equally important to investigate. Impact ofclimate changes are felt most strongly through changes in climate extremes. Anypositive or increasing trend in the extreme rainfall events is also a serious concern.The recent extreme heavy rainfall event occurred over Mumbai on 26 th July 2005prompts us to think whether there is any significant trend in extreme rainfall eventsover different parts of India.One of the most significant consequences of global warming due to increasein greenhouse gases would be an increase in magnitude and frequency of extremeprecipitation events. These increased extreme precipitation events can be attributedto increase in moisture levels, thunderstorm activities and large scale storm activity.In the global warming scenario, climate models generally predict an increase in largeprecipitation events (Houghton et al 2001). The numerical modelling community anddata analysts have shown interest on the issue of extreme events occurring aroundthe world.The recent studies have shown that there is an increasing trend of extremeprecipitation events in USA and Australia (Easterling et al. 2000, Haylock andNicholls 2000, Groisman et al. 2001; Kunkel 2003), western New Zealand(Salingerand Griffiths 2001), the UK in winter (Osborn et al. 2000), and south Africa(Fauchereau et al. 2003). Extreme rainfall events in Canada show no trend (Zhang2

et al. 2001; Kunkel 2003). Significantly decreasing trends in extreme rainfall eventshave been found in Western Australia (Haylock and Nicholls, 2000), south-east Asiaand parts of central Pacific (Griffiths et al. 2003), northern and eastern New Zealand(Salinger and Griffiths 2001), UK in summer (Osborn et al, 2000). Haylock et al.(2006) have recently addressed the trends in total and extreme rainfall over SouthAmerica and their links with sea surface temperatures.In India also, some studies have addressed this important issue. Rupa Kumaret al. (1992) examined the trends in the total precipitation during 1871-1984 andfound increasing trends in the precipitation amounts all along the west coast andnorthwest India. Their study also suggested a decreasing trend in the overallprecipitation in the eastern Madhya Pradesh. The study of Chhabra et al. (1997)indicates a decrease in the precipitation in hilly stations and an increase in theprecipitation in the urbanized/industrialized cities. Singh and Sontakke (2002)studied the fluctuations of precipitation amounts during 1829-1999 for the Indo-Gangetic Region. Their study indicates a significant trend from 1939 over the centralpart, and a significant decreasing trend over eastern parts of the country. Soman etal (1988) analysed annual extreme rainfall for the stations in Kerala state and foundthat stations in hilly terrain show a decreasing trends. Guhathakurta and Rajeevan(2006) analysed rainfall trends over 36 meteorological sub-divisions using a fixedrain-gauge network of over 1460 stations. Their study revealed significantdecreasing trends in rainfall over 3 meteorological sub-divisions (Jharkhand,Chattishgarh and Kerala) during the southwest monsoon season (June toSeptember).However, there are only a couple of studies on addressing the changes inextreme precipitation events. Sinha Ray and Srivastava (2000) examined the trendin the occurrence of heavy rainfall events in India. They analysed rainfall data of 151stations and considered a threshold of 7 cm and above. Sen Roy and Balling (2004)analysed the trends in the patterns of extreme precipitation events from 1910 to2000 and showed an increasing trend over most of western India including DeccanPlateau and a decreasing to a neutral trend over the eastern half of the countryexcept the northeastern corner. Sen Roy and Balling (2006) analysed the spatialpatterns of trends in the frequency and intensity of precipitation over India and3

concluded that most extreme events have become more frequent, particularly in thewestern half of the country. Francis and Gadgil (2006) using 37 years of rainfall dataexamined intense rainfall events over the west coast of India. The probability ofoccurrence of intense rainfall events is high from mid June to mid August. They haveanalysed the synoptic features associated with these intense rainfall events. KleinTank (2006) examined the changes in daily temperatures and precipitation extremesin central and south Asia. For this study, they have used daily data of 1961-2000.However, no robust signal of changes in precipitation extremes is observed over theregion. The only index with a significant (5% level) positive trend is the precipitationamount on very wet days. Also, the increase in the contribution of very wet days tothe total amounts between 1961 and 2000 is significant at 5% level, implyingdisproportionate changes of the precipitation extremes. Alexander et al (2005)examined global observed changes in daily climate extremes of temperature andprecipitation using a suite of climate change indices derived from daily data. Theyhave considered the data of 1951-2003 for the analysis. They have gridded theseasonal and annual climate change indices for the analysis. Their results indicate ageneral tendency towards wetter conditions throughout the 20 th century.There are many indices for examining the extreme rainfall events (Peterson etal. 2001). The earlier studies on extreme rainfall over India examined only a coupleof such indices. The joint working group on climate change detection of WorldMeteorological Organisation (WMO-CCL) and the research program on ClimateVariability and Prediction CLIVAR (Peterson et al., 2001) recommended 15 indiceson extreme rainfall. In this study, we have considered all these 15 indices andexamined the long term changes associated with these indices using 100 years ofdata. About half of the indices considered are expressions of anomalies relative tothe local climatology in the standard-normal period 1961-90 enabling comparisonsbetween stations in different countries and regions. We have considered daily dataof longer period (1901-2000) for the present analysis. The present study also dealswith analysis for the extreme rainfall during the southwest monsoon season as wellas annual rainfall over India. However, in this report, only the results of the analysisfor the southwest monsoon season (June to September) are discussed, which arefound similar with the annual rainfall data also.4

In section 2, details of data used and the methodology adopted are discussed.In section 3, the results are discussed in detail and in section 4 conclusions aredrawn.2. Data and MethodologyFor calculating extreme rainfall indices, we have considered daily rainfall datarecorded at rain-gauge stations over India. India Meteorological Department (IMD),maintains rainfall data observed at more than 6000 raingauge stations. However,only 199 observatories are maintained by the IMD personnel. Rest of theobservatories is manned by other state government agencies. For better quality, wehave considered only rainfall recorded at 199 IMD observatory stations. Daily rainfalldata of IMD stations for the period 1901-2000 are considered for the present study.Data for a year is considered to be missing if the daily rainfall data are not availablefor more than10% for the year. Filling of the missing daily data was not consideredas rainfall is a highly variable parameter. Indices of extremes are sensitive tochanges in station location, exposure, equipments and observing practices. Thedaily rainfall data archived at the National Data Centre (NDC) of IMD are qualitycontrolled and the outliers were identified. These outliers have been cross checkedwith the manuscripts and corrections were carried out whenever needed. Afterfiltering the data, we could get only 100 stations for the period 1901-2000, in whichmore than 90% of time daily data are available. It may be mentioned that Sinha Rayand Srivastava (2000) considered stations with data more than 60% of time, whichmay not be suitable for studies on extreme rainfall. Locations of these selectedstations are shown in Fig 1. Since we have put the condition of maximum 10%missing days for annual rainfall values, few stations are available in the eastern partof central India. Maximum stations are available from peninsular India.A list of over 50 internationally agreed climate change indices for temperatureand rainfall (WMO – CCl / CLIVAR) are available on web site (http://www.wmo.int)with their explanation and equations for calculating them. For the present study, wehave considered a set of 10 rainfall indices (Table -1). Detailed procedures tocalculate these indices are available at http://www.knmi.nl/samenw/eca.5

From the list prescribed by WMO, we have not considered indices of R10mmand R20mm, because rainfall events of 10 or 20 mm are quite common in the Indiansub-continent during the southwest monsoon season. Hence, we have consideredR7.5cm (75 mm) and R12.5cm (125mm) to study the trends in heavy precipitationfrequencies of rainfall greater than or equal to 7.5cm and 12.5cm.The definitions of indices allow seasonal and monthly partitions. As mentionedearlier, here, we have considered only the southwest monsoon season (June –September), the main rainy season in India. Most of the indices are defined in termsof counts of days crossing the thresholds either absolute (fixed) thresholds orpercentile (variable) thresholds. Annual or seasonal day-count indices based onpercentile thresholds are expressions of anomalies relative to the local climate.Consequently, the value of the thresholds is site specific. Such indices allow forspatial comparisons, because they sample same part of precipitation (probabilitydensity) distributions at each station. On the other hand, annual/seasonal day-countindices based on absolute thresholds are less suitable for spatial comparisons. Thereason being, over an area as large as India, thresholds of sample may vary differentfor different parts of the precipitation distributions. This means that in another climateregime the variability in such indices readily stems from another season. Forexample, R7.5cm may be an extreme event in North-west India, while such eventsare quite common along the west coast and NE India.Extreme rainfall indices considered for the analysis are calculated for eachyear for each station. Trends of these indices are calculated using linear regressionmodel to calculate the magnitude of the trends. Since most of the indices are countsof days, a non-parametric model will be the best suited to test the significance ratherthan using Students’ t test. Hence, we have used the Kendall -Tau test (Press et al.,1986) to examine the significance of the trend. The trends referred to as significantexceed the 95% or 99% confidence limits in the Kendall–Tau test. As mentionedearlier, missing data was not filled up while calculating the indices and trends.6

3. Results and DiscussionsWe have calculated the trends of 16 indices for monsoon season and onannual scale and estimated the statistical significance using the Kendal-Tau test. Forthe present study, results and discussions are presented for the monsoon season inrespect of the indices listed in Table-1. However, the trends with the annual data arealso found similar. Table – 2 shows the percentage of stations showing significanttrends in each index. The details of the trends in extreme indices are given below.Significant trend values are plotted in Figs 2-6 and magnitude of the trends perdecade are listed in Table-3.R7.5cm – Number of days with rainfall 7.5cm or moreFor the monsoon season, 63% of stations did not show any trend. 17% of thestations have shown positive significant trends. The significant positive trends areobserved (Fig. 2a) mainly along the west coast and some parts of central India andover Bihar. Higher values of trends are observed over stations of northern parts ofwest-coast. However, Mahabaleshwar has shown a significant negative trend innumber of days with 7.5cm or more.R12.5cm – Number of days with rainfall 12.5cm or moreWest coast stations and two stations of foothills of Himalayas have significantpositive trends (Fig. 2b) during the monsoon season. Similar to R7.5cm,Mahabaleshwar has shown a significant negative trend in number of days with12.5cm or more.RX1day – Highest one day total and RX5day – highest 5 day totalOut of the 100 stations, 73% of the stations during the monsoon season haveshown positive trend. However, the trends at only 15% of stations are statisticallysignificant. These stations (Fig. 2c) are mainly along the west coast of India, parts ofcentral India, northern parts of peninsular India and few stations in West Bengal and7

Bihar. Over these regions, one day total rainfall has shown an increasing trend.Similar results are also observed for the highest Five day precipitation (RX5day).The regions of significant positive trends (Fig. 2d) are mainly along the west coastand some parts of west central India. The magnitude of trends is also higher in theseareas.R75p – Moderate wet days – Days with rainfall more than 75 percentile of the longperiod average.Seventeen percent of stations have shown significant positive trends. Thesestations are distributed (Fig. 3a) along the west coast of India, northern parts of Indiaand west central India (west Madhya Pradesh). Only 5% of stations have shownsignificant negative trends.R75pTOT – Precipitation fraction due to moderate wet daysThe index R75pTOT determines the contribution of moderate wet days(rainfall with more than 75 percentile) on the seasonal rainfall total. During themonsoon season, 19% of stations showed significant positive trend. These stations(Fig. 3b) are mainly over along the west coast, southwest parts of central India,northern peninsular India, and few stations in Orissa. Over these regions, thecontribution from the rainfall amount higher than 75 percentile has increased. Only3% stations have shown significant negative trend.In a similar fashion, two more indices for 95 percentile and 99 percentile arealso calculated. Similar to R75p, the R95p index also shows significant positivetrends for 16% of stations. Significant positive trends are also noticed for 15% of thestations during the monsoon season for the index R95pTOT. These significanttrends (Fig. 3c & 3d) are along the west coast and northwestern parts of PeninsularIndia, which includes Maharashtra and parts of northern Karnataka. The indices ofR99p and R99pTOT showed (Fig. 4a and 4b) significant positive trends along thefoothills of Himalayas and south peninsular India.8

Regional Average TrendsIn the time series of station data, changes in routine observing practices mayhave introduced inhomogeneities of non-climatic origin that severely affect theextremes (Klein Tank et al. 2006). The results for individual stations may be affectedby inhomogeneities in the underlying series that were not detected. Therefore, robustconclusions can be drawn from tendencies over large areas and regional averagetrends. Moreover, changes in extreme precipitation due to global warming may bebetter revealed in regional average trends and not at the station level. This is due tothe fact that due to global warming, enhancement of moisture level may take placeon a larger scale and the associated enhancement in convection may take placeover a larger area. Therefore, we have repeated the analysis by dividing the countryinto 17 areas as shown in Fig. 5. Number of stations used for area averaging is alsoindicated in the regional boxes. The region-1 is the west coast with 10 stations.The results are shown in Table – 4. The changes observed at the stationsalong the west coast are realized even if we average the data over the west coast.The results from individual stations and area average data are consistent. Similarly,over the region-11 (northern parts of India), the signals are consistent. However,stations over the region-2 (south peninsula) have not shown much significantchanges in the climate indices. But in the area averaged analysis, many indices(RX1DAY, R75p, R75pTOT and R95p) have shown significant increasing trends.Similarly, individual stations over NE India (Region-14) have not shown anysignificant changes in the climate indices. However, when we made area averageover the region, many indices have shown significant trends.Fig.6. shows the time series of RX1DAY, RX5DAY, R95p and R95pTOTaveraged over the west coast (region). All the time series are showing an increasingtrend, especially the R95pTOT. During the recent years (after 1990), the per cent ofseasonal rainfall from very wet days (R95pTOT) is consistently more than 25%.9

6. ConclusionsFrom the present study, the following main conclusions can be drawn.1. Most of the extreme rainfall indices have shown significant positive trendsover the west coast stations and northwestern parts of Peninsula(Maharashtra). Along the west coast, contribution from the heaviest (95and 99 percentile) events to the total seasonal rainfall has increasedsignificantly. This result is important in view of the recent deluge occurredover Mumbai.2. Positive trends were also observed over the region 20 0 to 30 0 N and 75 0 to80 0 E.3. Individual stations over North east India did not show any significanttrends. However, averaged over the area, some climate indices(R75pTOT, R95p, R95pTOT) showed increasing trends.4. It is interesting to note that two hilly stations considered in this study(Simla and Mahabaleshwar) have shown significant decreasing trends ofsome extreme rainfall indices.Changes in the climate indices revealed in this study may be associated withvariability in the large-scale atmospheric circulation patterns. However, to whatextent the trends observed in this study are related with atmospheric circulationchanges needs to be addressed in future studies.The area of concern is northern parts of west coast and adjoiningMaharashtra. This study suggested that extreme rainfall events over this region hasshown a significant increasing trend. The study of Guhathakurta and Rajeevan(2006) has brought out the result that this region has also witnessed significantincrease in monsoon seasonal and annual rainfall. This inference was also madeusing about 100 years of rainfall data. Using a high resolution regional climatemodel, Rupa Kumar et al. (2006) have shown in the global warming scenario,summer monsoon precipitation over most parts of Maharashtra and adjoining southGujarat region is expected to increase on an average by 5 mm/day. This will10

translate into a huge increase in seasonal rainfall over this region. If we assume theobserved trend in increase of intense rainfall events will also persist into future, assuggested by the climate models, probability of intense rainfall events (like the oneexperienced at Mumbai in 2005) in near future will also increase proportionally.AcknowledgementsWe are thankful to Dr (Mrs) N.Jayanthi, LACD ADGM (Research) and Shri.Thakur Prasad, DDGM (Climatology) for providing facilities to do this work and forencouragement. We also thank Shri.G.S.Prakasa Rao, Director for his support forthis work. The first author participated in the WMO CLIVAR International Workshopon Enhancing South-central Asian Climate Change Monitoring and Indices held atPune in Feb 2005 and benefitted from the workshop.11

ReferencesAlexander, L. et al., 2005. Global observed changes in daily climate extremes oftemperature and precipitation. J. Geophys.Res., 32: D05109, doi10.1029/2005/JD006290.Chhabra, B. M., Prakasa Rao, G. S. and Joshi U. R., 1997. A comparative study ofdifferences in the averages of temperatures and rainfall over the Indianstations during the periods 1931 – 60 and 1961-90. Mausam 48: 1, 65-70.Easterling D R, Evans J L Groisman P Y, Karl T R, Kumbel K E, Ambenje P. 2000.Observed variability and trends in extreme climate events: a brief review.Bulletin of the American Meteorological Society 81: 417 – 425.Faucherean N. Trzasku S. Roualt M, Richard Y. 2003. Rainfall variability andchanges in the southern Africa during twentieth century in the global warmingcontext. Natural Hazards 29: 139 – 154.Francis, P.A. and Gadgil, S., 2006. Intense rainfall events over the west coast ofIndia. Meteorol Atmos. Phys., doi 10.1007/s00703-005-0167-2.Griffiths G M, Salinger M J, Leleu I. 2003. Trends in extreme daily rainfall across thesouth Pacific and relationship to the South Pacific convergence zone.International Journal of Climatology 23: 847 – 869.Groisman P Y, Knight R W, Karl T R. 2001. Heavy precipitation and high stream flowin the contiguous United States : Trends in the twentieth century. Bulletin ofthe American Meteorological Society 82: 219 – 246.Guhathakurta. P., and Rajeevan. M., 2006, Trends in the rainfall pattern over India,NCC Research Report No 2/2006, May 2006, India MeteorologicalDepartment, pp 23.Haylock M. R., Nicholls N, 2000. Trends in extreme rainfall indices for an updatedhigh quality data set for Australia 1910 – 1998. International Journal ofClimatology 20: 1533 – 1541.Haylock M. R. et al., 2006. Trends in total and extreme south American rainfall in1960-2000 and links with sea surface temperature. Journal of Climate19:1490-1512.12

Houghton Jr, Ding Y, Griggs Dj, Noguer M, Van der Linden PJ, Dai X, Maskell K,Johson CA (eds)., 2001. Climate Change 2001 ‘The Scientific Basis’,Cambridge University Press, Cambridge, UK.Klein Tank et al., 2006. Changes in daily temperature and precipitation extremes inCentral and South Asia. J.Geophys.Res., 111: No D16, D16105, doi10.1029/2005JD006316.Kunkel K E . 2003. North American trends in extreme precipitation. Natural Hazards29: 291 – 305.Osborn T J, Hulme M, Jones P D, Basnett T A. 2000. Observed trends in the dailyintensity of United Kingdom precipitation. International Journal of Climatology20: 347 – 364.Peterson T. C, Folland C. Gruza G, Hogg W, Mokssit A and Plummer, 2001. Reporton the activities of the working group on climate change detection and relatedrapporteurs 1998 – 2001. WMO Rep. WCDMP 47, WMO – TD 1071, Geneva,Switzerland 143 pp.Press W. H., Flannery B. P., Teukalsky, S. A., and Vetterling, W. T. 1986. NumericalRecipes: The art of scientific computing; Cambridge Univ., Press Cambridgepp 488 – 493.Rupa Kumar, K., Pant, G. B., Parthasarathy, B. and Sontakke, N., 1992. Spatial andsub-seasonal patterns of the long-term trends of Indian summer monsoonrainfall. Int. Jou. Clim.12: 257-268.Rupa Kumar, K., and co-authors, 2006, High resolution climate change scenarios forIndia for the 21 st century. Current Science. 90: 3, 334-345.Salinger M J, and Griffiths G M 2001. Trends in New Zealand daily temperature andrainfall extremes. International Journal of Climatology 2; 1437 – 1452.Shouraseni Sen Roy and Balling Jr R C. 2004; Trends in extreme daily precipitationon indices in India., Int. J. Climatol., 24: 457 – 466.Shouraseni Sen Roy and Balling Jr R C. 2006; Analysis of spatial patterns of trendsin the frequency and intensity of Indian precipitation. Mausam, 57: 3,431-436.Singh, N. and Sontakke, N. A., 2002, On climate fluctuations and environmentalchanges of the Indo Gangetic plains in India, Climate Change, 52: 287-313.Sinha Ray, K.C. and Srivastava, A. K. 2000, Is there any change in extreme eventslike heavy rainfall? Current Science, 79: 2, 155 – 158.13

Soman M K, Krishnakumar K, Singh N. 1988. Decreasing trend in the rainfall ofKerala. Current Science 57: 5 – 12.Zang X, Hogg W D, Mekis F. 2001. Spatial and temporal characteristics of heavyprecipitation events over Canada. Journal of Climatology 14: 1923 – 1936.14

Table – 1Extreme Rainfall Indices considered for the studyS. N. Description Code1. Heavy Precipitation days (Precipitation ≥ 7.5 cm) R7.5cm2. Very Heavy Precipitation Days (Precipitation ≥ 12.5 cm) R12.5cm3. Highest 1 day precipitation amount RX1day4. Highest 5 day precipitation amount RX5day5.Moderate wet days(days with RR > 75 th percentile of daily precipitation R75pamount)6. Precipitation fraction due to Moderate wet Days R75pTOT7. Very wet days (days with RR > 95 th percentile) R95p8. Precipitation fraction due to Very wet days R95pTOT9. Extremely wet days (days with RR > 99 th percentile) R99p10. Precipitation fraction due to extremely wet days R99pTOTTable -2Per cent of stations showing significant trendsR7.5cmR12.5cmRX1DAYIndex % No.of StationsHaving +ve trend 32 11 73 78 57 70 42 75 13 70-ve trend 5 2 26 22 22 22 6 21 3 28+ve Sig. trend 17 8 15 9 17 19 16 17 10 15-ve Sig. trend 2 1 1 3 5 3 1 1 3 1RX5DAYR75PR75PTOTR95PR95PTOTR99PR99PTOT15

Table - 3Significant trend values (per decade) for the rainfall indicesStationLongLatR7.5cmR12.5cmRX1dayRX5dayR75pR75pTOTR95pR95pTOTR99pR99pTOTAGRA 78.03 27.17 0.40AHMEDABAD AERO 72.63 23.07 1.40AJMER 74.62 26.45 6.60 0.10 1.30ALIPUR 88.33 22.53 0.50AMBALA 76.77 30.38 0.10 0.10 3.30 5.90 0.40 1.20 0.10 0.10 0.80AMRAOTI 77.78 20.93 0.10 1.00ANGUL 85.10 20.83 0.10 0.90 0.20 1.30BAHRAICH 81.60 27.57 0.10 8.70 0.10 0.80BALASORE 86.93 21.50 0.10 4.70 0.80 0.10 0.90BANGALORE C.O. 77.58 12.97 0.20 0.10BEGUMPET/HYD.AER 78.47 17.45 2.80 0.70BELGAUM 74.53 15.85 0.60BIJAPUR 75.72 16.82 1.80 0.20BOMBAY COLABA 72.82 18.90 0.30 0.10 8.00 12.10 0.60 1.00 0.20 0.10 0.60DARBHANGA 85.90 26.17 -8.00 -0.70 -1.40 -0.10GAYA AERODROME 84.95 24.75 -0.10 -3.60 -5.20 -1.20 -0.10 -1.30HOSHANGABAD 77.77 22.77 0.30INDORE 75.80 22.72 0.10 3.80 10.50 0.30 0.90 0.20 1.60JAGDALPUR 82.03 19.08 -0.50JALPAIGURI 88.72 26.53 9.10 0.50 1.00 0.10 1.10KAKINADA 82.23 16.95 0.10KHANDWA 76.37 21.83 0.10 5.50 0.30 0.80 0.10 1.40KOCHI FORT 76.23 9.97 0.50 1.00KURNOOL 78.07 15.83 2.60 0.30 1.60 0.20 1.60LUDHIANA 75.87 30.93 0.20MAHABALESHWAR 73.67 17.93 -0.70 -0.40 -0.70 -0.70 -0.20MAINPURI 79.05 27.23 0.10 0.40 1.00 0.10MANGALORE 74.85 12.87 0.30 0.10 4.60 0.80 0.20 0.80 0.10 0.60NELLORE 79.98 14.45 0.40NEW DELHI/SAFDAR 77.20 28.58 0.40NUNGAMBAKKAM 80.25 13.07 0.50PANJIM 73.82 15.48 0.30 0.20 7.60 13.80 0.80 0.30 1.20 0.10 0.70PARBHANI 76.77 19.27 0.10 0.90PHALODI 72.37 27.13 -0.20PORT BLAIR 92.72 11.67 0.10 0.40PURI 85.82 19.80 0.10 0.80 0.10 1.20PURNEA 87.47 25.77 0.20 0.10 5.40 1.10 0.10 0.80RAICHUR 77.35 16.20 0.10 1.10 0.80RATNAGIRI 73.33 16.98 0.50 0.20 6.30 11.00 0.80 1.10 0.40 1.50 0.10 0.60SIMLA 77.17 31.10 -5.70 -0.60 -0.60 -0.10SRINAGAR 74.83 34.08 1.10SURAT 72.83 21.20 0.20 3.50 10.20 0.90TEZPUR 92.78 26.62 0.40 0.10 0.90TIRUCHI.PALLI(A) 78.72 10.77 1.40VERAVAL 70.37 20.90 0.2 0.10 6.90 17.00 0.50 1.70 0.20 2.20VISAKHAPATNAM AP 83.23 17.72 0.3016

Table - 4Rainfall Indices (1901-2000) – Significant TrendsIndexR7.5cmR12.5cmRX1DAYRX5DAYR75pR75pTOTR95pR95pTOTR99pR99pTOTRegion1 + ve N + ve + ve N + ve + ve + ve + ve + ve2 N N + ve N + ve + ve + ve N N N3 N N N N N N N N N N4 N N N + ve N N N N N N5 N N N N N N N N N + ve6 N N N N N N N + ve N N7 N N N + ve N N + ve N + ve N8 N N N N N N N N N N9 N N N N N N N N N N10 N N N N N N N N N N11 N N N N N + ve + ve + ve N N12 N + ve N + ve N N N N + ve + ve13 N N N N N N N + ve N N14 N N N N N + ve + ve + ve N N15 N N N ─ ve ─ ve N N N N N16 N N N N N N N N N N17 + ve N N N N + ve N N N NN : No Significant Trend + ve : Positive Significant Trend─ ve : Negative Significant Trend17

Fig. 1 : Stations considered for the study18

abcd- Significant Positive Trends - Significant Negative TrendsFig. 2 : Significant trends of rainfall indices (a) R7.5cm (b) R12.5cm (c) RX1day and(d) RX5day for Southwest Monsoon Season (June to September).19

abcd- Significant Positive Trends - Significant Negative TrendsFig. 3 : Significant trends of rainfall indices (a) R75p (b) R75pTOT(c) R95p and(d) R95pTOT for the Southwest Monsoon Season (June to September).20

ab- Significant Positive Trends - Significant Negative TrendsFig. 4 : Significant trends of rainfall indices (a) R99p (b) R99pTOT for thesouthwest Monsoon season (June to September).21

Figures in the parenthesis indicate the number of stations averaged in the regionRegion 1 : West coast with 10 stations. Region 2 & 3 are without the west coast stations.Fig. 5 : Regions considered for the study22

Rainfall in mm200150100500Rainfall Index RX1day for Region 1Trend = 2.43 mm / decade1901191119211931194119511961197119811991YearRainfall in mm450400350300250200150100500Rainfall Index RX5day for Region 1Trend = 5.98 mm / decade190119111921193119411951Year1961197119811991Fig 6: Time series of the Extreme Rainfall Indices for the region 123

Percentage50403020100Rainfall Index R95pTOT for Region 1Trends = 1 percent / decade190119111921193119411951Year1961197119811991Rainfall Index R95p for Region 120Trend = 0.3 days / decade15Days1050190119111921193119411951Year1961197119811991Fig 6 (contd) : Time series of the Extreme Rainfall Indices for the region 124

N C C RESEARCH REPORTS New statistical models for long range forecasting of southwest monsoon rainfall over India,M. Rajeevan, D. S. Pai and Anil Kumar Rohilla, Sept. 2005. Trends in the rainfall pattern over India, P. Guhathakurta and M. Rajeevan, May 200625