2015 5th International Conference on Environment Science and Engineering Volume 83 of IPCBEE (2015) DOI: 10.7763/IPCBEE. 2015. V83. 22 Statistical Analysis of Long Term Temporal Trends of Precipitation and Temperature in Wainganga Sub-Basin, India Arun Kumar Taxak 1, A. R. Murumkar 1 and D. S. Arya 1 1 Department of Hydrology, IIT Roorkee, 247667 Uttarakhand, India Abstract. Present study explores changes in precipitation, and mean, maximum & minimum temperatures of seven stations located in the Wainganga sub-basin, Central India. Annual trends were analyzed using Mann-Kendall test and Sen's slope estimator test for two time periods 1901-2002, and 1971-2002 to study recent changes in last 30 years. All stations are exhibited increasing trends in mean, maximum, minimum temperatures and potential evapotranspiration; decreasing trends are found in precipitation and wet day s frequency during both the periods. Comparison of trends during these periods depicts that the trends have become strong during the last 30 years except maximum temperature. Keywords: Climate change, Wainganga basin, Trend, Temperature, Mann-Kendall Test, Sen's slope, Wavelet analysis 1. Introduction Studies of climate change and climate change impacts on various sectors require urgent attention as it might lead to increased risk of hunger and water scarcity, rapid melting of glaciers, and decreased river flows etc. [1]. The latest report of the Intergovernmental Panel on Climate Change (IPCC) identifies an increase in the average global surface temperature of approximately 0.6 0 C since the late 19th century with an increase in the rate of temperature rise in the most recent decades [2]. Several studies have reported that trends of warming were not uniform throughout the day; less warming in maximum temperatures and substantial warming in minimum temperatures is reported [3][4]. Some studies have shown that the trend and magnitude of warming over India/the Indian sub-continent over the last century is broadly consistent with the global trend and magnitude [5][6]. Consolidation of recent studies on analysis of rainfall which is the key input into the hydrologic system shows conclusive evidence that rainfall is decreasing in India [6]-[8]. In India precipitation fluctuations are largely random with no systematic change detectable [9]. Changes in evaporation have been given little attention barring a few studies. In the above context, the present study is an attempt to determine annual trends and interrelationship among mean (T mean ), maximum (T max ), minimum (T min ) temperatures; and, precipitation, wet day s frequency and potential evapotranspiration (PET) for Wainganga sub-basin of India. 2. Study Area and Data Used 2.1.Study area Wainganga sub-basin is part of the Godavari River basin located in central India from 78 0 00' to 80 0 53' East longitudes and 19 0 40' to 22 0 41' North latitudes as shown in Fig. 1. The total catchment area of the basin is 51,421 km 2 spread over three states Maharashtra, Madhya Pradesh and small portions of Chhattisgarh states. The elevation ranges from 144 to 1208 m (Fig. 1) above mean sea level in the basin. The river in its Corresponding author. Tel.: +918265998554 E-mail address: aruntaxak@gmail.com 129
initial reaches flows westwards and thereafter turns southwards in Madhya Pradesh and continues to flow Southwards through the Maharashtra State. Nagpur is the biggest city in the basin. Fig. 1: Location of Wainganga basin and Stations 2.2. Details of data Monthly data of mean, maximum and minimum temperature, potential evapo-transpiration (PET), precipitation and wet-day frequency data of seven stations lying in the basin were downloaded from "http://www.indiawaterportal.org/metadata" for the last century (1901-2002). Location of stations is shown in Fig. 1. Data quality control is a necessary step; however, since Mann Kendall (MK) test is a rank based non-parametric test, they are robust against outliners. As all series are complete no gap filling was required. The monthly data was averaged to provide annual values for each year and the analyzed for monotonic trends. 3. Methodology All annual series are checked for auto-correlation by using student s t-test at 5% significance level. The non-parametric test, Mann Kendall (MK)/Modified Mann-Kendall (MMK) test is applied for non-auto correlated/auto correlated annual series for monotonic trend detection. Sen s slope estimator test is used to detect the magnitude of change over time. Change in percentage is discussed in terms of percentage change over mean for 100 years. 30 year window is adopted to show recent trends as indicated by several authors that for climate purpose minimum 30 year window should be adopted. It should be noted that different results could potentially be found using different start and end years to define the recent years. Tests are applied for both the periods (1) 1901-2002 for long term changes and (2) 1971-2002 for recent changes. Correlation coefficient between PET and temperature and precipitation was evaluated to show interrelationship between them. The wavelet analysis is applied to check any oscillations in climatic variables [10]. 4. Results and Discussion 130
The results of MK/ MMK test and Sen s slope estimator test of annual series of climatic variables at seven stations for 1901-2002 and 1971-2002 are listed in Table 1 and Table 2, respectively. Significant autocorrelation is found in annual series of PET, mean, maximum and minimum temperature, therefore Modified Mann Kendall (MMK) test is applied for these series. Precipitation and wet day s series are free of auto-correlation; therefore they are subjected to MK test. Increasing trends are observed for temperature (mean, maximum and minimum) and potential evapotranspiration while decreasing trend is observed for precipitation and wet day frequency at all stations for both the study periods. Significant decreasing trend is observed in precipitation at four stations at 5% significance level in the basin for 1901-2002 periods. Among temperature variables minimum temperature is showing more increase than maximum temperature. Decreasing trend is observed in both the precipitation and wet day frequency implies decrease in precipitation due to lesser rainy days. Table 1: Table showing MK (MMK) z value and Sen s slope test value for 1901-2002 Stations T Mean T Max T Min PET Prep Wet day z slope z slope z slope z slope z slope z slope Balaghat 1.36* 0.006 1.26* 0.006 1.45* 0.006 0.41* 0.002-2.174-1.610-1.758-0.030 Bhandara 1.38* 0.006 1.32* 0.006 1.36* 0.007 0.63* 0.003-2.238-2.018-1.047-0.020 chhindwara 1.29* 0.006 1.26* 0.006 1.28* 0.006 0.56* 0.003-1.052-0.882-1.064-0.020 Gharchiroli 1.74* 0.009 1.69* 0.009 1.79* 0.009 3.24 0.004-0.590-0.417 0.584 0.009 Gondhiya 1.44* 0.007 1.34* 0.006 1.47* 0.007 0.60* 0.003-2.481-2.007-1.515-0.027 Nagpur 1.29* 0.006 1.26* 0.006 1.30* 0.006 0.59* 0.003-1.949-1.424-0.711-0.012 Seoni 1.33* 0.006 1.24* 0.006 1.38* 0.006 0.45* 0.003-1.116-0.865-1.220-0.021 *indicates presence of autocorrelation and bold indicate significant trend Table 2: Table showing MK (MMK) z value and Sen s slope test value for 1971-2002 Stations T Mean T Max T Min PET Prep Wet day z slope z slope z slope z slope z slope z slope Balaghat 1.038 0.008 0.542 0.006 2.030 0.010 0.775 0.016-1.379-3.922-1.534-0.162 Bhandara 1.193 0.006 0.604 0.005 1.131 0.008 0.387 0.005-0.170-0.412-0.697-0.093 chhindwara 1.627 0.011 1.317 0.009 1.844 0.012 0.589 0.008-1.410-4.997-1.131-0.124 Gharchiroli 1.596 0.010 1.317 0.008 1.937 0.013 0.527 0.008 0.139 0.399 0.325 0.026 Gondhiya 1.038 0.007 0.356 0.004 1.751 0.011 0.604 0.008-0.201-0.572-1.007-0.121 Nagpur 1.317 0.007 0.914 0.007 1.503 0.009 0.077 0.002-0.201-0.590-0.418-0.067 Seoni 1.131 0.008 0.666 0.005 1.580 0.008 0.465 0.010-0.883-3.935-1.224-0.139 Table 3 shows the trend analysis for whole basin for both the periods. The increasing trend is observed for temperature and evapotranspiration and decreasing trend for precipitation and wet days for 102 years periods. Similar trends are observed during the recent period (1971-2002). No significant trend is observed for any variable for both periods for whole basin. Results show that trends are stronger in the recent years as % change during 1971-02 period is more than the 1901-02 periods. The changes in temperature in recent years are more attributed to minimum temperature than the maximum temperature resulting in overall more change in recent years. The overall change (% change/100 years) in precipitation in the basin was -11.77% during the 1901-02 period and -14.58 % for 1901-2002 period. The overall change (% change/100 years) in mean, maximum and minimum temperature was 2.44oC, 1.92oC and 3.23oC for 1901-2002 respectively. The overall change (% change/100 years) for 1971-2002 in mean, maximum and minimum temperature was 3.02oC, 1.81oC and 5.23oC for 1901-2002 respectively. Table 3: Table showing mean, MK ( z value) and % change over 100year for 1901-2002 and 1971-2002 Variables 1901-2002 1971-2002 Mean Z value % Change Mean Z value % Change T Max 32.95 1.36* 1.92 33.17 0.76 1.81 T Min 20.30 1.48* 3.23 20.54 1.84 5.28 T Mean 26.61 1.44* 2.44 26.83 1.49 3.02 PET 78.47 0.63* 0.38 78.49 0.56 0.79 Prep 1209-1.897-11.77 1155-0.64-14.58 Wet 58-1.041-3.23 57-0.91-17.69 * indicates presence of autocorrelation 131
Analyzing the correlation of PET with precipitation and temperature, as shown in Table 4, it is found that PET is negatively correlated with precipitation (-0.38) and positively with temperature. In other words the maximum correlation is found with maximum temperature with correlation coefficient 0.58 for 1901-2002 and 0.70 for 1970-2002. The correlation of PET with temperature increases with time. Table 4: Correlation of PET with precipitation, temperature 1901-2002 1971-2002 P rep T Mean T Max T Min P rep T Mean T Max T Min PET -0.38 0.50 0.58 0.41-0.37 0.53 0.70 0.32 Fig. 2: Showing wavelet power spectrum for (a) T Max, (b)t Mean (c)t Min, (d) PET, (e) Precipitation and (f)wet day Wavelet analysis is a method to investigate signals that change over time. Dark regions indicate periods in the data, where it oscillates with the corresponding frequency. Morlet wavelet is most widely used in climate research studies. In order to verify the trends, wavelet analysis is used. It gives the relative power at a certain scale and a certain time. Wavelet transforms of annual series of precipitation, maximum temperature, minimum temperature, mean temperature, PET and wet days for whole basin is shown in Fig. 2. Wavelet analysis of precipitation, maximum temperature, minimum temperature, mean temperature, PET and wet days is found that power of wavelet spectrum is low during the study period. As high power is concentrated around 100 years, which is equal to the study period, therefore trends are more global. Results show that 132
there is no oscillation in the time series and the trend can be attributed to anthropogenic climate change (global warming). 5. Conclusions The trends analysis of annual series of temperature (mean, maximum and minimum), PET, precipitation and wet days s of Wainganga basin, India for 1901-2002 and 1971-2002 periods for long term and recent changes were carried out. Analysis shows a broad general pattern of increasing trend of temperature and PET and decreasing trend in precipitation and number of wet days in the basin. All temperatures and PET shows an increasing trend in either series. PET was found negatively correlated with precipitation and positively correlated with temperature. Observed decreasing trend in both the precipitation and wet day s frequency and increasing trend in PET implies more dry conditions in the basin. As the LULC of the area in the basin is mostly agricultural land and forests, the increase in PET and decrease in precipitation will have a significant impact on economy and ecosystem of the region. 6. References [1] IPCC, 2007. Summary for policymakers. In Climate Change 2007: The Physical Science Basis, Soloman S, Qin D, Manning M, Chen Z, Marquis M, Ayeryt KB, Tignor M, Miller HL (eds). Intergovernmental Panal of Climate Change, Cambridge University Press: UK. [2] Houghton, J. T., Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden, X. Dai, K. Maskell, and C. A. Johnson (Eds.) (2001), Climate Change 2001: The Scientific Basis, Cambridge Univ. Press, New York. [3] Dai, A., A. D. Del Genio, and I. Y. Fung (1997), Clouds, precipitation and temperature range, Nature, 386, 665-666. [4] Easterling, D. R.,et al. (1997), Maximum and minimum temperature trends for the globe, Science, 277, 364-367. [5] Arora, M., Goel, N. K. & Singh, P. (2005) Evaluation of temperature trends over India. Hydrol. Sci. J. 50(1), 81 93. [6] Dash, S. K., Jenamani, R. K., Kalsi, S. R. and Panda, S. K. (2007) Some evidence of climate change in twentiethcentury India. Climatic Change 85, 299 321. [7] Goswami, B. N., Venugopal, V., Sengupta, D., Madhusoodanam, M. S. & Xavier, P. K. (2006) Increasing trends of extreme rain events over India in a warming environment. Science 314, 1442 1445. [8] Basistha A., Arya D. S., Goel N.K. (2009) Analysis of historical changes in the Indian Himalayas. International Journal of Climatology. 29: 555-572 (2009). [9] Lal, M. Nozawa, T.; Emori, S,;Harasawa, HH.; Takahashi, K.; Kimoto, M.; Abeouchi, A., Nakajima, T.; Takemura, T., Numaguti, A. 2001Future climatic change: Implications for Indian summer monsoon and its variability. Current Science, 81, n (9) p 1196-207. [10] Torrence, C., and G. P. Compo, 1998: A practical guide to wavelet analysis. Bull. Amer. Meteor. Soc., 79, 61 78. 133