MAUSAM, 68, 1 (January 2017), 51-66

551.571.1 (540.63)

Some characteristics of southwest monsoon rainfall over urban entres in and Telangana

K. NAGA RATNA and MANORAMA MOHANTY* Meteorological Centre, Hyderabad, *Meteorological Centre, Ahmedabad, India (Received 4 November 2015, Accepted 28 July 2016) e mail : [email protected]

सार – इस शोध पत्र म तटीय आ ध्र प्रदं ेश (CAP) के तटीय क्षेत्र के िलए शहरी टेशन और आसपास के टेशन तथा तेलंगना एवं रायलसीमा के 46 वष (1969-2014) के दैिनक वषार् आँकड़ का चयन िकया गया है। इन टेशन के समची ू अविध के समायण, मानक िवचलन एवं प्रसरण गणाु ंक, टी टे ट उपयोग करत हए मह े ु वपणू परीक्षणर् , मान-कै डेल टे ट के आकड़ को िलया गया हँ ै। इन टेशन का चयन समान अविध वाले आकड़ कँ े आधार पर िकया गया है। इस प्रकार सभी टेशन पर िकए गए टी-टे ट से ऋतिन ु ठ वषार् (JJAS) म स भी टेशन पर साथक पिरणाम र् (P < 0.001) देखने को िमले ह। इसक े अलावा मेन-कै डेल टे ट का उपयोग करत हए जे ु ेड-आँकड़ प्रा त िकए गए ह जो े आंध्र ेश प्रद रा य के तटीय टेशन ग नावरम, मछलीपटनम और िवशाखापटनम के आकड़ ँ की सटीकता म 95 प्रितशत की साथक विध को दशार् ृ र्ते ह। त ेलंगना, अंदनी टेहशन ैदराबाद (शहरी कद्र ) म अ यिधक भारी वषार् की घटनाओ म ं 90 प्रितशत तर तक मह वपणू विध पाई गई हर् ृ ै। इसके बाद प्र येक शहरी कद्र (िवशाखापटनम, ग नावरम, मछलीपटनम तथा हैदराबाद) के अलग-अलग अ ययन िकए गए और प्रा त पिरणाम से पता चला िक शहरी क्षेत्र के कद्र पर आस -पास के अ य टेशन की तलना म वषाु र् म मह वपणू विध पाई गई हर् ृ ै। तटीय आध्र प्रदं ेश के तटवतीर् टेशन पर तेलंगना एवं रायलसीमा के अंदनी टेशन की तलना म वषाु र् की मात्रा म मह वपणू वर् िध दृ ेखी गई है।

ABSTRACT. In the present study daily rainfall data for 46 years (1969-2014) was selected for the urban stations and surrounding stations for coastal areas of Coastal Andhra Pradesh (CAP) and inland areas of Telanagana (TEL) and Rayalaseema (RSM). The statistics such as regression, standard deviation and coefficient of variance, significance test using t-test, Mann-Kandell test were worked out for the entire period for the stations. The stations were selected on the basis where the period of data is same. The t-test thus performed for all stations showed significance (p < 0.001) in seasonal rainfall (JJAS) for all the stations. Further z-statistics using Mann-Kandell test was performed that showed significant increase at 95% confidence level for Gannavaram, and Visakhapatnam along the coast of Andhra Pradesh state. Over Telengana, Hyderabad (Urban centre) an inland station, showed significant increase at 90% level of confidence for extreme heavy rainfall events. Henceforth, seperate studies for each urban centre (Visakhapatnam, Gannavaram, Machilipatnam and Hyderabad) were done and results showed significant increase in rainfall over urban centres compared to other surrounding stations and the significant increase in rainfall was observed for the coastal stations along Andhra Pradesh coast when compared to inland stations of Telanagana and Rayalaseema.

Key words – Seasonal rainfall, Rainydays, Coastal stations, Inland stations, Man-Kandell test.

1. Introduction climate of the state. Andhra Pradesh and Telanagana have three meteorological sub-divisions namely, Coastal Andhra Pradesh bordered by Telangana in the north Andhra Pradesh (CAP), Rayalaseema (RSM) and lies between 12°41' and 22° N latitude and 77° and Telangana (TEL) as defined by India Meteorological 84°40' E longitude with Bay of Bengal in the East. Department (IMD) 1999. CAP consists of nine districts Among the other states, which are situated on the and this again subdivided to North Coastal Andhra country's coastal area, Andhra Pradesh has got a coastline Pradesh (NCAP) and South Coastal Andhra Pradesh of around 972 km, which gives it the 2nd longest coastline (SCAP) for easy representation of climate parameters in the nation. Southwest monsoon during July and according to their climatic sub-regions, which comprises continues till September has major role in determining the of Srikakulam, Visakhapatnam, Vizianagaram, East

(51)

52 MAUSAM, 68, 1 (January 2017)

Godavari, West Godavari (NCAP) and Krishna, , surface level and easterlies in upper level (100 hPa). Prakasam, Nellore (SCAP) respectively. The climate of Extreme heavy rainfall events cause intense rainfall events Andhra Pradesh is generally classified as sub-humid and that lead to severe floods and landslides causing damage dry (over NCAP) and wet-wetter and semi-arid (over to property and life and influence the economical status of SCAP), which receive total rainfall about 1128 mm and the state. Therefore climatology studies also have 996 mm respectively. Rayalaseema is a geographic region importance for understanding the climate changes in the state of Andhra Pradesh in India. It occupies atleast happening during long term period over a specified 42% of the state territory, with TEL to the north and the location or region for well planned management and CAP region of Andhra Pradesh to the east. The region is growth of the state economy. divided into southern zone (Chittoor and Kadapa) and scarce rainfall zone (Anantapur and Kurnool) which are In this global warming era, the monsoon variability classified as semi-arid and arid zone. Telangana had been challenging to the scientific community. Many comprises of North Telangana (Adilabad, Nizamabad, studies have attempted to determine the trend in rainfall Krimnagar, Warangal and Khammam) and South on both large and regional scales. Most of these deal with Telangana (Medak, Nalgonda and Ranga Reddy). South the analysis of annual and seasonal series of rainfall for Telanagana (STEL) is semi-arid and arid area and has a some individual stations or group of stations. predominantly hot and dry climate. North Telangana Subbaramayya and Naidu (1997) have examined the trend (NTEL) is semi arid (wetter in some districts). in rainfall for different sub-divisions and the whole of India were examined for the period 1871-1988. Rajeevan As a result, the regions are marked by large scale et al. (2006) analysed rainfall series using 1476 rainguage variations in land characteristics, vegetation and lead to stations data for the period 1901-2003. Mohanty et al. complex circulations embedded with large scale monsoon (2014) analysed the rainfall for different stations of flow. The rainfall extremes during monsoon season leads Gujarat for the period 1969-2010 and found that there is to unusual floods and drought over the regions. The significant increase trend of rainfall over coastal areas of permanent and semi permanent synoptic features over Gujarat. Studies of extreme rainfall trends in India Indian sub-continent in the large-scale monsoon showed that increase in frequency of intense rainfall circulation, causes spatial and temporal variability in the events lead to decrease in number of moderate rainfall rainfall distribution. The geography and land surface play events and total seasonal and annual rainfall. Also vital role, influencing convective activity and rainfall Rajeevan et al. (2008) showed that increased trend of intensity of the sub-divisions. The total annual rainfall extreme rainfall events could be associated with increased varies from 566 mm over Anantapur in Rayalaseema trend in sea surface temperature and surface heat flux. (RSM) sub-division to 1135 mm over Kakinada in Coastal Goswami et al. (2006) showed that there was significant Andhra Pradesh (CAP). The total annual rainfall varies increasing trend in the frequency and magnitude of from 649 mm over Nalgonda to 1149 mm of rainfall over extreme rainfall events and significant decreasing trend of Ramagundam in Telangana (TEL). Moreover, southwest moderate events over central India during monsoon monsoon comprises nearly 2/3rd of the total annual season, leading no significant trend in mean rainfall. Also rainfall. The southwest monsoon advances in first week of some studies, particularly over hilly terrain over Kerala June and covers all the three sub-divisions by second indicated decreasing trend in extreme annual rainfall for week of June. July and August are most rainy months some stations and non-significant increasing trend for contributing 25%-35% of the annual rainfall play vital most of the stations. Aim of the study is to understand the role in agriculture production and economy of the two concept of urbanization and their impacts on urbanization states. The rainfall occurs due to the influence of the to help the economic planning and economic synoptic scale systems like monsoon trough/low and developments of the states. Part-2 includes the data and depressions that usually develop over Bay over Bengal. methodology used for the purpose in the present study. These systems bring moisture supply to the coast and Part-3 gives details of the results and further discussion of some parts of the inland areas. On an average there were the climate variables like rainfall, rainydays and extreme 10-12 rainy days and 12-18 rainy days over Andhra rainfall events. Pradesh and Telangana respectively. The withdrawal of monsoon begins during first week of October. During 2. Data and methodology monsoon season, the two states receive heavy (HRF), very heavy (VHRF) and extremely heavy rainfall (EHRF) in 2.1. Data association with cyclonic circulations such as monsoon lows/depressions that develop over Bay of Bengal and To find out the characteristic features of the south- move westwards/northwestwards/westnorthwards or due west monsoon season rainfall over subdivisions of CAP, to monsoon troughs with predominant westerlies at RSM and TEL (Fig. 1), daily data for the stations over a

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 53

20.00 standard deviation and coefficient of variance, significance test using Mann-Kandell test were worked 19.00 out for the entire period (1969-2014) for the stations. NIZAMABADRAMGUNDAM KALINGAPATNAM 18.00 The Mann-Kendall test is a non-parameteric test for KHAMMAM VISAKHAPATNAM HYDERABAD identifying trends in time series data. The test was BHADRACHALAM 17.00 KAKINADA suggested by Mann (1945) and has been extensively used MAHABUBNAGAR GANNAVARAM with environmental time series (Hipel and McLeod, MACHILIPATNAM 16.00 2005). KURNOOL NANADYALA ONGOLE

15.00 Let X1, X2, X3, …, Xn represents n data points where ANANTAPUR Xj represents the data point at time j. Then the Mann- KADAPA NELLORE SUBDIVISIONS Kendall statistic (S) is given by: 14.00 TELANAGANA AROGYAVARAM COASTAL ANDHRA PRADESH Sxxjnk sign , 2,3,4... and  1,2,3,...j 1 13.00   jk RAYALASEEMA 12.00 where, 76.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 85.00

Fig. 1. Locations of sub-divisions and stations of Andhra Pradesh and sign XX 1, if XX0 Telangana state  jkjk

=0, ifXXjk0

period of 46 years (1969-2014) was considered for the study. These rainfall data were obtained from National =-1, ifXXjk0 data centre (NDC) Pune, India and Meteorological Centre, Hyderabad, India. The stations considered for the present A very high value of S is an indicator of an study are Kalingapatnam (18.33/84.13), Visakhapatnam increasing trend, and a very low negative value indicates a (17.68/83.3), Kakinada (16.95/82.23), Gannavaram decreasing trend. For the sample size >30, a normal (16.53/80.8), Machilipatnam (16.18/81.13), Ongole approximations to the Mann-Kendall test may be used. (15.57/80.05), Nellore (14.45/79.98), Kunool For this variance S is obtained as, (15.8/78.07), Nandyala (15.47/78.48), Anantapur (14.68/77.62), Kadapa (14.48/78.83), Arogyavaram nn 12 n 5 tptp 12 tp  5 (13.53/78.5) situated in Andhra Pradesh state and VS   Hyderabad (17.45/78.47), Nizamabad (18.67/78.1), 18 Ramgundam (18.77/79.43), Mahabubnagar (16.75/78), pq 1, 2, 3, ... Bhadrachalam (17.25/80.15) and Khammam (17.67/80.88) situated in Telagana state. where, tp is the number of ties for the pth value and q is the number of tied values. 2.2. Methodology Then, standard statistical is computed by: The daily rainfall was averaged over the months (June, July, August and September) and (JJAS) monsoon ZS1 ifS  0 season as a whole. Average rainy days and heavy rainfall VS events for different monsoon months and season as a whole were also computed and analysed. The month-wise and annual frequency of heavy, very heavy and extremely S 1 ifS < 0 heavy rainfall events were found out. For the present VS study, daily 24 hr accumulated rainfall events were considered for the study period of 46 years (1969-2014). If the ‘z’ score is positive or negative value indicates As per IMD terminology of IMD heavy (64.5 mm to increasing or decreasing trend of the total population 124.4 mm), very heavy (124.5 mm to 244.4 mm) and respectively. If calculated value is equal to or greater than extremely heavy rainfall (>244.4 mm) were considered in the table value (1.65, 1.96 and 2.58), the trend is this study. The month-wise comparison and interannual significant at a particular level of significance (10%, 5% variability was analysed. The statistics like regression, and 1% respectively).

54 MAUSAM, 68, 1 (January 2017)

(a) (b) June CV%/Rainfall(mm) July CV%/Rainfall(mm) 19.00 51.0 19.00 46.8 50.7 RMD 47.7 RMD NZB NZB 149.6 163.4 77.2 289.9 304.9 58.3 KLN KLN 132.5 150.8 18.00 18.00 58.0 77.6 57.8 51.9 KMT WLT KMT WLT 47.6 119.1 51.0 126.4 HYD 62.7 132.2 HYD 70.8 255.8 108.4 BDC 163.9 BDC 117.9 54.8 245.6 53.4 17.00 55.3 KND 17.00 57.5 KND MBN 52.6 126.1 MBN 52.9 166.9 100.0 GNV 153.4 GNV 117.266.7 193.549.8 MPT MPT 16.00 87.6 99.1 16.00 59.2 182.8 KRN 84.7 KRN 73.1 97.1108.3 ONG 120.965.3 ONG NDL 61.1 NDL 99.5 82.3 148.6 15.00 15.00 84.5 99.7 ANT 92.7 131.5 ANT 59.4 63.3 56.2 CDP NLR 68.6 CDP NLR 76.0 45.4 119.2 84.3 14.00 14.00 93.3 67.3 ARV ARV 71.7 87.4 13.00 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

(c) (d) August CV%/Rainfall(mm) September CV%/Rainfall(mm) 19.00 48.3 19.00 58.9 55.9 RMD 76.6 RMD NZB NZB 284.3 277.2 58.7 159.6 175.9 62.6 KLN KLN 162.2 184.3 18.00 18.00 55.5 57.3 69.5 61.2 KMT WLT KMT WLT 50.0 159.1 61.2 185.3 HYD 65.2 237.3 HYD 75.6 163.9 197.8 BDC 142.5 BDC 222.3 49.8 125.1 72.9 17.00 50.9 KND 17.00 60.4 KND MBN 45.6 171.8 MBN 61.9 180.7 188.5 GNV 150.3 GNV 177.951.3 162.166.4 MPT MPT 16.00 54.5 170.3 16.00 60.2 168.8 KRN 67.9 KRN 69.7 146.183.9 ONG 142.369.6 ONG NDL 110.8 NDL 135.9 167.2 138.2 15.00 15.00 94.8 64.7 ANT 68.2 66.2 ANT 61.2 62.4 84.6 CDP NLR 132.2 CDP NLR 117.0 90.3 140.1 87.0 14.00 14.00 61.3 57.3 ARV ARV 106.7 133.9 13.00 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

(e) JJAS CV%/Rainfall(mm) 19.00 27.1 30.9 RMD NZB 883.3 921.4 29.5 KLN 629.8 18.00 42.0 27.6 KMT WLT 26.0 589.9 HYD 53.0 789.2 612.6 BDC 711.0 31.3 17.00 34.7 KND MBN 34.2 645.5 592.1 GNV 650.635.0 MPT 16.00 33.7 621.0 KRN 39.1 506.546.8 ONG NDL 536.3 407.2 15.00 68.6 ANT 40.3 40.5 341.6 CDP NLR 452.4 307.0 14.00 30.6 ARV 399.7 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

Figs. 2(a-e). Average rainfall (mm) and coefficient of variation(%) during June, July, August and September months and season for 46 years (1969-2014). CV is mentioned above and Rainfall below

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 55

(a) (b) June rainydays July rainydays 19.00 8.4 19.00 13.7 8.6 RMD 13.7 RMD NZB NZB 6.3 8.6 KLN KLN 18.00 18.00 7.3 5.8 12.8 8.2 7.0 KMT WLT 9.6 KMT WLT HYD 6.7 HYD 11.6 BDC BDC 7.0 10.7 17.00 6.7 KND 17.00 10.8 KND MBN 6.8 MBN 11.8 GNV GNV 6.0 10.4 MPT MPT 16.00 5.3 16.00 7.8 KRN 4.1 KRN 5.9 5.0 ONG 8.4 ONG NDL NDL

15.00 15.00 3.2 4.2 ANT 4.2 3.2 ANT 6.8 5.9 CDP NLR CDP NLR

14.00 14.00 4.2 6.3 ARV ARV

13.00 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

(c) (d) August rainydays September rainydays 19.00 12.9 12.9 RMD 19.00 8.2 NZB 7.7 RMD 9.3 NZB KLN 8.4 KLN 18.00 8.5 18.00 11.8 9.0 KMT WLT 7.8 11.0 KMT WLT HYD 7.6 11.3 HYD BDC 7.1 10.3 BDC 17.00 KND 8.8 11.0 17.00 KND MBN 8.2 10.7 MBN GNV 8.5 GNV 9.7 8.4 MPT MPT 16.00 8.9 16.00 7.9 KRN 7.1 8.5 KRN 6.8 ONG 7.3 ONG NDL NDL

15.00 15.00 5.0 6.8 ANT 6.9 5.9 ANT 7.1 5.1 CDP NLR CDP NLR

14.00 14.00 6.2 7.1 ARV ARV

13.00 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

(e) Season JJAS rainydays 19.00 10.8 10.7 RMD NZB 8.2 KLN 18.00 9.9 7.9 8.8 KMT WLT HYD 9.2 BDC 9.2 17.00 9.2 KND MBN 9.5 GNV 8.6 MPT 16.00 7.5 KRN 6.0 7.3 ONG NDL

15.00 4.8 ANT 6.3 5.0 CDP NLR

14.00 5.9 ARV

13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

Figs. 3(a-e). Average rainydays during June, July, August and September months and season for 46 years (1969-2014)

56 MAUSAM, 68, 1 (January 2017)

The standard deviation is calculated by using the rainfall over Telangana (TEL) was found be maximum formula given below: during July with Ramgundam (304.9 mm) and Nizamabad (289.9 mm); August with Nizamabad (284.8 mm) and 1 N 2 Ramagundam (277.2 mm). It was also noted that Coastal  x x Andhra Pradesh received higher rainfall in July with N i i1 Gannavaram (193.1 mm) and Machilipatnam (182.8 mm); August with Gannavaram (177.1 mm) and Machilipatnam σ = standard deviation (170.8 mm). The rainfall over Rayalaseema was least as compared to the other sub-divisions. Nandyala received xi = each value of dataset maximum rainfall during July and August with 148.6 mm and 167.2 mm respectively. Rayalaseema received more x = the arithmetic mean of the data rainfall during July, August and less during September (This symbol will be indicated as and June respectively. Anantapur is the supposed to be mean from now) station that receives scarce rainfall than any other station. Average rainfall recorded was 56.2 mm, 68.6 mm, N = the total number of data points 84.6 mm and 132.2 mm respectively in the months of June, July, August and September. However minimum 2 2 rainfall was received over Nellore, Anantapur and Ongole x  x = The sum of x  x for all data  i i  during June. As such, it is noticed that there was points considerable increase in rainfall in all the regions with well establishment of monsoon circulation as season progress from July, August and September. The population CV can be estimated using the ratio of the sample standard deviation σ to the sample mean x : CAP recorded higher rainfall as compared to RSM and North Telangana (NTEL) recorded higher rainfall as compared to South Telangana (STEL). Season as a whole  CV  very high rainfall was recorded over Gannavaram x (650 mm) in Andhra Pradesh and Ramagundam (921.4 mm) in Telanagana. Very less rainfall was 3. Results and discussion recorded over Nellore (307 mm) in Andhra Pradesh and Hyderabad (612.6 mm) in Telangana. The t-test performed for the selected stations showed significance (p < 0.01) in seasonal rainfall (JJAS) for all 3.2. Rainy days (rainfall events greater than 2.5 mm) the stations. The analysis for coefficient of variance, average rainfall, rainydays (>2.5 mm) and extreme rainfall The number of rainy days (days with 24 hours events (> 64.5 mm) was presented in section 3.1, 3.2 and cumulative rainfall (>2.5 mm) for June, July, August, 3.3 respectively. Interannual variability and trend analysis September and season as a whole are presented in of the climate parameters rainfall, rainydays and extreme Figs. 3(a-e). The rainy days are found to be maximum in rainfall events was discussed in section 3.4. number (15-25 mm/day) during the peak months July and August months followed by September and June. TEL 3.1. Rainfall variability recorded maximum number of rainy days during July and August as compared to June and September. The rainy The average monthly and seasonal rainfall and their days are more in number during July, August and coefficients of variance (CV) of the stations of Andhra September for the CAP and RSM, whereas less number in Pradesh and Telangana state were shown in Figs. 2(a-e). June. Further, TEL was found to have more rainy days The average rainfall was found to be high during July, than CAP and RSM. The frequency of rainy days and the August and September when compared to June. The rainfall amounts during monsoon season showed the analysis of coefficient of variance for each station progression and intensity of the monsoon season over the indicated rainfall variability is more during June as region. However, this represented the magnitude of compared to July, August and September. The rainfall convective activity over Andhra Pradesh and Telangana vaiariability is noticed to be more over RSM when states. The time series of the subdivisions were compared to CAP and TEL. This showed as moving north presented in Fig. 7. This showed rainfall increased and to south the rainfall variability increased. The rainfall rainy days decreased over a period of time, which intensity was found to decrease from north to south over revealed the intensity of rainfall increased over the three Andhra Pradesh and also Telangana states. The average sub-divisions.

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 57

(a) (b) June Heavy rainfall events July Heavy rainfall events 19.00 0.44 19.00 0.67 0.20 RMD 0.02 0.65 RMD 0.11 NZB NZB 0.06 0.20 0.28 KLN 0.06 KLN 0.04 18.00 18.00 0.17 0.29 0.54 0.14 0.20 KMT WLT 0.02 0.30 KMT 0.02 WLT HYD 0.15 HYD 0.02 0.34 BDC BDC 0.24 0.11 0.11 17.00 0.11 KND 0.02 17.00 0.15 KND 0.02 MBN 0.22 MBN 0.20 0.02 GNV 0.02 GNV 0.04 0.09 0.15 MPT 0.02 MPT 0.04 16.00 0.24 16.00 0.24 KRN 0.06 KRN 0.06 0.020.02 ONG 0.28 ONG NDL 0.02 0.02 NDL 0.04

15.00 15.00 0.09 0.11 ANT 0.09 0.07 0.06 ANT 0.11 0.15 0.02 CDP 0.04 NLR 0.04 CDP NLR

14.00 14.00 0.06 0.09 ARV 0.02 ARV

13.00 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

(c) (d) August Heavy rainfall events September Heavy rainfall events 19.00 0.63 19.00 0.54 RMD 0.09 0.26 NZB 0.11 0.44 RMD 0.09 0.02 0.28 NZB 0.04 KLN 0.02 0.39 KLN 0.06 18.00 0.02 0.48 0.36 18.00 KMT WLT 0.07 0.22 0.50 0.46 0.04 KMT WLT 0.07 HYD 0.41 0.04 0.04 0.36 HYD BDC 0.06 0.02 0.19 0.22 BDC 17.00 KND 0.28 0.28 0.02 17.00 KND MBN 0.24 0.06 0.02 0.37 MBN 0.02 0.04 0.02 GNV 0.54 0.00 GNV 0.07 0.22 0.24 MPT 0.02 MPT 0.09 16.00 0.35 16.00 KRN 0.41 0.21 0.17 KRN 0.28 ONG 0.02 0.26 ONG NDL NDL 0.04

15.00 15.00 0.11 0.17 ANT 0.11 0.19 0.09 ANT 0.17 0.26 0.15 CDP NLR CDP 0.02 NLR 0.02

14.00 14.00 0.06 1.13 ARV ARV

13.00 13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

(e) Season JJAS Heavy rainfall events 19.00 2.00 1.83 RMD 0.31 NZB 0.22 0.02 1.15 KLN 0.19 0.02 18.00 1.41 1.29 1.37 KMT 0.11 WLT 0.17 HYD 0.09 1.04 BDC 0.17 0.85 17.00 0.78 KND 0.13 MBN 0.04 1.33 0.06 0.02 GNV 0.13 0.69 MPT 0.17 16.00 1.24 KRN 1.24 0.59 0.020.77 ONG NDL 0.77 0.13

15.00 0.48 ANT 0.48 0.68 0.33 CDP 0.68 NLR 0.06

14.00 1.35 ARV 1.35

13.00

77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00

Figs. 4(a-e). The monthly and seasonal frequencies of daily 24 hr cumulative heavy, very heavy, extremely heavy rainfall during 46 years (1969-2014) over the stations of Andhra Pradesh and Telangana

58 MAUSAM, 68, 1 (January 2017)

(a) (b)

(c) (d)

(e) (f)

(g)

Figs. 5(a-g). The time series of the seasonal rainfall over urban centres Machilipatnam, Gannavaram, Kalingapatnam, Visakhapatnam, Anantapur, Kunool and Hyderabad stations

3.3. Heavy rainfall events (greater than 64.5 mm) presented in Figs. 4(a-e). It is observed that the frequency of heavy rainfall events (HRF) is very high during July, The monthly and seasonal frequencies of daily 24 hr August, September and followed by June. The frequency cumulative heavy, very heavy, extremely heavy rainfall of very heavy rainfall (VHRF), extreme heavy rainfall during past 46 years (1969-2014) over the stations of events (EHRF) is high during the months of July, August, Andhra Pradesh and Telangana states were analysed and September and followed by June. The frequency of HRF,

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 59

(a) (b)

(c) (d)

(e) (f)

(g)

Figs. 6(a-g). The time series of the seasonal rainydays over urban centres Machilipatnam, Gannavaram, Kalingapatnam, Visakhapatnam, Anantapur, Kurnool and Hyderabad stations

VHRF and EHRF is more during the months of July, of HRF and VHRF is less and EHRF are found only in a August and September as compared to June month. It may few occasions during all the monsoon months, over a few be due to the impact of monsoon lows/depressions that stations except September where the EHRF was form over Bay of Bengal and move westwards or received along coast line. This may be due to the northwestwards. It was also observed that the frequency reason that monsoon lows/depression frequency cross the

60 MAUSAM, 68, 1 (January 2017)

Figs. 7(a-c). The time series of the seasonal rainfall and rainydays over (a) Coastal Andhra Pradesh, (b) Telangana and (c) Rayalaseema

Andhra Pradesh coast. The time series of the seasonal 3.4. Trend analysis of rainfall rainydays over urban centres Machilipatnam, Gannavaram, Kalingapatnam, Visakhapatnam, Anantapur, The Mann-Kandell trend test (z-statistics) was done Kurnool and Hyderabad stations of Andhra Pradesh and for the period of 46 years (1969-2014) for the rainfall Telangana as given in Figs. 6(a-g) were analysed. It was tendency over all the stations for the season as a whole seen that the frequency of heavy rains was high over and each month of the season. Thus the z-scores Nizamabad (2.1/year), Ramagundam (2.3/year), computed for the seasonal means of rainfall for inland Khammam (1.5/year) and Hyderabad (1.4/year) in stations of TEL/RSM and coastal stations in CAP are Telangana. It was recorded for Gannavaram (1.5/year), presented in Table 1. Among these stations, three stations Visakhapatnam (1.5/year), Machilipatnam (0.9/year) and noted to be robust, that indicated significant increasing Kalingapatnam (1.4/year) in Andhra Pradesh. It was trend of 95% level of significance with z score of 1.97* found that EHRF events found to be more in NTEL for Gannavaram, 2.12* for Machilipatnam and 1.97* for followed by CAP. However in STEL, EHRF was more Visakhapatnam. All the three stations located were in over Hyderabad (1.4/year) as compared to its surrounding Coastal Andhra Pradesh along the coast. Gannavaram and stations. Machilipatnam located in SCAP () and

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 61

TABLE 1 may be clearly noted that increased trend in seasonal rainfall over major cities (Hyderabad) and developing Z-statistics of seasonal rainfall for selected stations during the years 1969-2014 cities (Visakhapatnam, Gannavaram, Machilipatnam) may be due to the impact of urbanization. Thus, the data is Rainfall Rainy days Rainfall > 64.5 mm considered separately and study was done for Time series Test Z Test Z Test Z understanding the urbanization. The present work is to aim for the impacts of urbanization of the cities as JJASknd 0.64 0.41 -0.31 compared to the surrounding locations. However, only the JJASkln 1.17 1.46 0.09 stations that showed significant increasing trend (above JJASmpt 2.12* 0.54 1.00 90% level of confidence) of rainfall and extreme heavy JJASnlr 0.68 -0.13 0.06 rainfall events were considered for detailed study. JJASong 0.18 -0.78 -0.67 Further the trend analyses also revealed that the data JJASgnv 1.97* 1.13 1.29 analysis performed for 1969-2014 for extremely heavy JJASwlt 1.97* 2.88* 0.31 rainfall (EHRF) events found to be well marked for Hyderabad with significant increase at 95% level of JJASCAP 1.69+ 1.08 0.26 confidence at z value of 1.71* during the monsoon JJASant 1.53 1.53 0.52 season as whole during the 46 years (1969-2014) period JJASarv 0.55 0.23 0.56 (Table 1). One of the reasons that such a phenomenon of JJAScdp 0.48 0.66 -0.03 increase rainfall could occur may be because an increase JJASkrn 1.44 0.70 1.14 in temperature (heat islands) that increases the capacity of JJASndl 0.26 0.30 -1.34 the atmosphere to hold water which in turn increases JJASRSM 1.02 0.86 0.28 the amount of precipitation. Also Mohapatra et al. (2009 and 2010) revealed that the increased rainfall may be due JJASbdc 0.68 -1.41 -0.41 to the impact of urbanization as found for other major + JJAShyd 0.89 -0.13 1.71 cities like Bangalore and Mumbai. The annual frequency JJASkmt 1.22 0.37 1.26 of the extreme heavy rainfall events showing significant JJASmbn -0.05 0.26 -1.28 increase for Machilipatnam, Gannavaram, Visakhapatnam JJASnzb -0.15 0.33 -0.84 and Hyderabad were shown in Figs. 8(a-d). JJASrmd -0.42 0.26 -1.42 JJASTEL 0.44 -0.49 -1.09 The surrounding stations showed different trends for EHRF events in TEL state. These stations were Ramagundam well known for its mordernisation and industrialization (Thermal power stations and coal mines) Visakhapatnam located in NCAP (Visakhapatnam and Nizamabad that were surrounded with hills, forest and district). The time series of the three sub-divisions are river showed insignificant decrease in Seasonal rainfall presented in Figs. 7(a-c). It is also noted that CAP and EHRF events. In RSM, Nandyala which surrounded showed significant increasing trend in rainfall at 90% by Nallamala Hills, dense forest to east and granite mines level of confidence (with z-score of 1.69+). Rainfall towards south and river to its west showed insignificant significantly increased at the rate of 2.37 mm/year (CAP), decrease in EHRF events. If this trend continues in the 1.93 mm/year (RSM), 0.492 mm/year (TEL) and inversely future then it could have repercussions in the proportional as rainydays decreased for all the three sustainability of surface water resources and groundwater subdivisions at the rate of 0.059/year (CAP), 0.045/year recharge. (RSM) and -0.048/year (TEL) respectively. It was also seen that during 1969-2014, there were 7 However, Hyderabad also exhibited insignificant monsoon depressions that formed over BOB and crossed increasing trend (z-score = 0.89) located in Telanagana. from sea to land between 12° N to 18° N latitude over Kurnool and Anantapur which is local urban centre in Andhra Pradesh coast. Out of these 7 monsoon Rayalaseema also exhibited insignificant increasing trend depressions (1969-2007); 5 crossed near to Machilipatnam in rainfall as compared to other cities near to its location. and Gannavaram and moved over land towards NTEL. Rest of the stations were selected, in order to test the While the other two monsoon depressions crossed above urbanization impact showed mixed trend with 17° N latitude near to Visakhapatnam. After monsoon insignificant decreasing or increasing trend in rainfall season of 2007 no such systems were observed to cross (Please refer to z-score values in Table 1. The time series the coast (IMD, 2008). This may be reason that impacts of of the seasonal rainfall and rainydays of the urban cities decreased trend of extreme heavy rainfall events over were presented in Figs. 4(a-e) and 5(a-g). Henceforth, it NTEL which is located inland. This would also be the

62 MAUSAM, 68, 1 (January 2017)

Figs. 8(a-d). The time series of the annual frequency of total heavy rainfall events (> 64.5 mm) showing significant increase over urban cities (a) Gannavaram, (b) Visakhapatnam, (c) Machilipatnam and (d) Hyderabad

Figs. 9(a-d). The time series of the seasonal rainfall events showing significant increase over (a) actual rainfall of Visakhapatnam and Visakhapatnam AP (b) decadal mean of actual rainfall of Visakhapatnam and Visakhapatnam AP (c) difference of actual rainfall of Visakhapatnam and Visakhapatnam AP and (d) difference of decadal mean of actual rainfall of Visakhapatnam and Visakhapatnam AP at 90% significance level

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 63

Figs. 10(a-e). The time series of the Seasonal rainfall events showing significant increase over (a) actual rainfall of Machilipatnam and Gannavaram (b) decadal mean of actual rainfall of Machilipatnam and Gannavaram (c) actual rainfall of Rentachintala (d) difference of actual rainfall of Machilipatnam and Gannavaram and (e) difference of decadal mean of actual rainfall of Machilipatnam and Gannavaram at 90% significance level

reason for occurrence of more moderate rainfall events the areas of formation of monsoon depressions (2.5 mm to 64.4 mm) than extreme heavy rainfall events continuously and during this period the vertical shear of over Visakhapatnam (in NCAP) during the study period. the horizontal wind between the lower and upper The reason for this may be convection becomes weaker troposphere is found to decrease. Naidu et al. (2011) over Bay of Bengal and its surrounding stations during the showed the prominent weakening of upper level easterlies recent years in this global warming era. Since 1965, in this global warming era and significant decrease satellites have been monitoring the weather systems for in southwest monsoon rainfall. Some results indicate that

64 MAUSAM, 68, 1 (January 2017)

Fig. 11. Periodicities of urban centres of coastal Andhra Pradesh

NAGA RATNA and MOHANTY : CHARACTERISTICS OF SW MONSOON R/F OVER URBAN CENTRES 65

although the number of low pressures has been increasing (Fig. 11). Considering the periodicity in difference of during the past decades, the dynamic conditions such as rainfall for Gannavaram and Rentachintala, no such wind shears, moisture, mean sea level pressure were not natural periodicity was seen. This shows that the increase favourable for the intensification to depressions and in rainfall over the city centres was not a part of natural cyclonic systems (Dash et al., 2004). variability, but due to urbanization impact.

Separate studies were done for examining the impact 4. Conclusions of urbanization and rainfall over Visakhapatnam and Gannavaram, which exhibited significant increasing trend The analysis of rainfall over the urban centres and its in rainfall. Inorder to understand the urbanization of surrounding locations during monsoon season (June Visakhapatnam, the rainfall was compared to the local through September) based on 46 years data revealed. Airport station located in the outskirts of the city, 15 kms away from the current station. The rainfall trend of the (i) The average rainfall was found to be maximum in two stations is shown in Figs. 6(a-g). The rainfall showed July and August for stations of Telangana followed by gradual raise in rainfall after 1991. It continued to Coastal Andhra Pradesh and Rayalaseema. The rainfall increase at the rate of 4.09 mm/year, while airport variability was found to be maximum during July and reported 2.01 mm/year. The difference in rainfall was August followed September and June. found to significantly increase at 90% level of confidence (z-score = 1.86*). The actual, 10 year mean rainfall and (ii) The frequency of heavy rainfall events was found to difference rainfall of Visakhapatnam and Airport were be high as compared to very heavy rainfall and extreme presented in Figs. 9(a-d). The rate of increase was faster heavy rainfall events and frequency was more during July, for Visakhapatnam city until 2010. After 2010, the August and September. Extreme heavy rainfall events Airport station also recorded the almost same rainfall were few in number. values, which thus revealed that impact of extension of urbanization into the outskirts of the city. Surprisingly, (iii) The time series of seasonal rainfall indicated over the Visakhapatnam city the rainfall and rainy days significant increase (95% level of significance) in seasonal increased significantly with z-score of 2.88* (95% level of rainfall for the stations of Gannavaram, Machilipatnam confidence) which showed increase in convective activity and Visakhapatnam located along the coastline in CAP. was very prominent during the monsoon with lower The seasonal rainfall found to be significant increase for rainfall amounts (2.5 mm to <64.5 mm). This may be due coastal area stations when compared to inland area to maritime air near to the coast causing rainfall due to stations. land and sea breeze effect. (iv) It was revealed that the CAP rainfall increased Gannavaram and Machilipatnam recorded significant significantly (90% level of confidence) and rainy days increase trend in rainfall as compared to the surrounding (>2.5 mm) decreased. Also, Telangana and Rayalaseema stations in Coastal Andhra Pradesh. This may be also due showed increasing and decreasing trend for seasonal to the impact of urbanization. The rainfall data was rainfall and rainy days respectively. The number of rainy compared with Rentachintala located in remote area near days less as compared to rainfall amounts indicates to Gannavaram. Rentachintala recorded decreasing rainfall intensity increased and more over coastal stations tendency in rainfall. The difference of rainfall with as compared to inland stations.

Gannavaram was well marked that showed significant (v) Over Visakhapatnam city (Coastal station) increasing trend in rainfall with z-score of 1.87+ (90% the rainfall and rainydays increased significantly level of confidence) and 2.06* (95% level of confidence) (95% level of confidence) which showed increase in respectively. The actual, 10 year mean rainfall and convective activity was very prominent during the difference rainfall of Gannavaram, Machilipatnam and monsoon with high lower rainfall events and less extreme Rentachintala are presented in Figs. 10(a-e). It was rainfall events. understood due to urbanization, rainfall was increasing over years. The reason may be due to unusual heat of the (vi) Over Hyderabad city (Inland station), the extreme surface, the hot surface can hold more water vapour and rainfall events significantly increased (90% level of hence precipitation takes place. To examine, whether the confidence), but the lower rainfall events are very less. increasing trend of the rainfall over city centres was inherent natural periodicity in rainfall, the power spectrum ( vii) It was also observed that significant increasing analyses of the rainfall data (1969-2014) was carried out. (decreasing) of contribution of extreme rainfall events It indicates that there were periodicities of 3 years and 9 may be balanced by significant decreasing (increasing) of years in case of all the stations except Gannavaram lower rainfall events.

66 MAUSAM, 68, 1 (January 2017)

(viii) However, density of station observations I ndia Meteorological Department, 2008, “Cyclone e-Atlas, electronic would be necessary to understand the chief features form of tracks of cyclones and depressions over north Indian Ocean, 1891-2014”, Chennai. of the variability of climate for specified region or location. Mann, H. B., 1945, “Nonparametric tests against trend”, Econometrica, 13, 245-259. Acknowledgements Manorama, Mohanty., Mohapatra, M. and Jaaffrey, S. N. A., 2014, Authors are thankful to the Director General of “Some Characteristics of Rainfall over Major Urban Centres of Meteorology, New Delhi, Deputy Director General of Gujarat”, Mausam, 65, 4, 608-618. Meteorology, Chennai and Director, M. C. Hyderabad for extending their support and providing the facilities to Mohapatra, M., Kumar, Naresh and Bandypadhyay, B. K., 2009, “Role of mesoscale low and urbanization on exceptionally heavy carry out the work. Also thanks are due to G. Krishna rainfall events on 26th July, 2005 over Mumbai : Some Kumar, NDC Pune for providing the data. observational evidences”, Mausam, 60, 3, 317-324.

Mohapatra, M., Kumar, Naresh and Bandyopadhyay, B. K., 2010, References “Unprecedented rainfall over Bangalore city during October 2005”, Mausam, 61, 1, 105-112.

Dash, S. K., Kumar, Rajendra, J. and Shekhar, M. S., 2004, “On the Naidu, C. V., DurgaLakshmi, K., Satyanarayana, G. Ch., Malleswara decreasing frequency of monsoon depressions over the Indian Rao, L., Ramakrishna, S. V. S. S., Rama Mohan, J. and Naga region”, Curr. sci., 86, 10, 1401-1411. Ratna, K., 2011, “An observational evidence of climate change during global warming era”, Global and Planetary change, 79, Goswami, B. N., Venugopal, V., Senugupta, D., Madhusoodanan, M. S. 11-19. and Xavier, K. Prince, 2006, “Increasing trend of Extreme rain events over India in a Warming Environment”, Science, 314, Rajeevan, M., Bhate, J. and Jaswal, A. K., 2008, “Analysis of variability 1442-1445. and trends of extreme rainfall events over India using 104 years of gridded daily rainfall data”, Geol. Res. Let., 53, 1-6. Hipel, K. W. and McLeod, A. I., 2005, “Time series Modelling of Water Resources and environmental Systems”, http://www.stats.uwo. Rajeevan, M. Bhate, Kale, K. D. and Lal, B., 2006, “High resolution ca/faculty/aim/1994Book. daily gridded rainfall data for the Indian region. Analysis of break and active monsoon spells”, Curr. Sci., 91, 292-306. India Meteorological Department, 1999, “Climatological tables of observatories in India”, Published by India Meteorological Subbaramayya, I. and Naidu, C. V., 1992, “Spatial variations and trends Department, Shivajinagar Pune. in the Indian monsoon rainfall”, Int. J. climatol., 12, 597-609.