4.1 Temporal Analysis
Annual
The temporal trends in annual rainfall and temperature show wide variations. First step was to apply different statistical measures like CS and CK. Both CS and CK are measures of data flatness, unevenness, distortion, symmetry, or more precisely, lack of symmetry, and their values for normal data distribution are close to zero. In the study area, Honavar and Panjim stations have high CK, while Thrissur and Dahanu have low CK. The stations around southern Kerala (Kumarakom, Painavu, Ponmudi, and Kovalam) only have negative CS which suggests that rainfall is decreasing over the southern part of Kerala and increasing in the rest of the study area. It is also found from the analysis that 1998, 2010, and 2019 were the warmest years in most of the stations in the study area, and 1991, 1994, 1999, and 2015 were the coldest years (Table 1). According to Hare (2003), CV is used to classify the degree of variability of rainfall events as low (CV < 20), moderate (20 < CV < 30), and high (CV > 30). Analysis of the last 40 years of annual total rainfall data from 27 representative ground-based meteorological stations in WG indicated a coefficient of variation ranging from 17.27 to 26.91 percent (highest = Dahanu, lowest = Udupi). Of the 27 stations considered, 12 stations had coefficient of variation above 20%, highlighting the extreme variability of rainfall over the study area (Table 1). In accordance with the above-cited paper, it is found that the study area lies in the moderate variability zone. It is also found that the CV value in the study area decrease from north to south, i.e., variability is high in the northern part of WG and low in the southern part. Apart from that, Fig. 2 depicts the linear rainfall trends at each station. The annual trends of all stations are positive except for Bandipur and Ratnagiri. The inter-annual variations in the temporal dataset are high (SD = 345.13, CV = 15.49), varying from 1456.63 mm (in the year 1986) to 2890.72 mm (in the year 2010), and the average rainfall is about 2228.01 mm (Fig. 2, Table 2). The highest rainfall in the year 2010 and the lowest rainfall in the year 1986 are attributed to a strong La-Nina and a moderate El-Nino, respectively. Tables 3 and 4 clearly show that the El-Nino and La-Nina phenomena have coincided with only two rainfall deficit years (1986, 1987) and two rainfall excess years (2007, 2010). Hence, we can say that rainfall along the WG is not affected by ENSO events but rather by local relief, cascading topography, and the length and width of mountain summits (Tawde and Singh 2015). Furthermore, rainfall over the study area followed a cyclic pattern of a 5-year moving average under changing climatic conditions (Fig. 3a). In the case of heavy rainfall events, we take the classification from an already existing study by Tawde and Singh (2015). A heavy rainfall event is thus defined as a heavy downpour (in mm) in a day ranging between 120 mm and 150 mm, with rainfall exceeding 150 mm/day classified as a very heavy rainfall event. Details of the events are shown in Fig. 3b. Thus, it is clearly seen in the graph that both events and their intensity increased after 2000 (Devika and Pillai 2020), and most of the events occurred around 19o N, i.e., the area around the Mumbai metropolitan region, and around 14o N to 16o N, i.e., the area around Panjim, Honavar, Dandeli, and Vengurla, the mining and quarrying regions of Goa (Gadgil 2011) (Fig. 3b).
Time
|
M
(mm)
|
SD
(mm)
|
CV
(%)
|
CK
|
CS
|
% of annual
|
Rainfall
|
R
|
Temperature
|
|
|
|
|
|
|
|
Z
|
Q
|
|
Z
|
Q
|
January
|
5.86
|
8.08
|
137.99
|
4.65
|
2.17
|
0.26
|
-0.55
|
-0.017
|
0.06
|
0.36
|
0.0027
|
February
|
7.67
|
10.79
|
140.73
|
2.86
|
1.88
|
0.34
|
0.71
|
0.029
|
0.00
|
2.46*
|
0.0148
|
March
|
17.02
|
25.42
|
149.32
|
27.39
|
4.86
|
0.76
|
1.50
|
0.215
|
0.04
|
2.18*
|
0.0137
|
April
|
47.45
|
28.65
|
60.38
|
0.18
|
0.87
|
2.13
|
0.80
|
0.340
|
0.00
|
2.20*
|
0.0139
|
May
|
121.29
|
85.43
|
70.43
|
4.57
|
1.90
|
5.44
|
1.53
|
1.231
|
0.03
|
1.43
|
0.0113
|
June
|
520.64
|
117.96
|
22.66
|
0.50
|
0.49
|
23.37
|
-0.90
|
-1.630
|
0.02
|
3.37α
|
0.0180
|
July
|
565.18
|
162.50
|
28.75
|
-0.49
|
-0.31
|
25.37
|
0.71
|
2.186
|
0.03
|
3.37α
|
0.0175
|
August
|
412.33
|
120.80
|
29.30
|
0.51
|
0.60
|
18.51
|
0.94
|
1.874
|
0.05
|
4.23α
|
0.0212
|
September
|
242.88
|
119.39
|
49.16
|
-0.97
|
0.46
|
10.90
|
2.13*
|
3.867*
|
0.11
|
4.63α
|
0.0192
|
October
|
190.62
|
76.65
|
40.21
|
0.09
|
0.53
|
8.55
|
1.15
|
0.969
|
0.08
|
3.25∞
|
0.0178
|
November
|
76.67
|
40.91
|
53.35
|
1.37
|
1.30
|
3.44
|
1.20
|
0.504
|
0.02
|
2.62∞
|
0.0184
|
December
|
20.57
|
18.06
|
87.78
|
1.02
|
1.24
|
0.92
|
0.97
|
0.207
|
0.04
|
1.25
|
0.0107
|
Winter (January-February)
|
13.53
|
13.82
|
102.16
|
0.98
|
1.36
|
0.61
|
0.59
|
0.069
|
0.01
|
1.67+
|
0.0092
|
Pre-Monsoon (March-May)
|
185.77
|
88.87
|
47.84
|
2.84
|
1.28
|
8.34
|
1.50
|
1.386
|
0.06
|
2.44*
|
0.0114
|
Monsoon (June-September)
|
1741.03
|
317.06
|
18.21
|
-0.85
|
0.09
|
78.14
|
1.36
|
7.541
|
0.06
|
4.25α
|
0.0186
|
Post-Monsoon (October-December)
|
287.86
|
95.17
|
33.06
|
0.05
|
0.53
|
12.92
|
1.85+
|
1.975+
|
0.08
|
2.83∞
|
0.0163
|
Annual
|
2228.01
|
345.13
|
15.49
|
-0.51
|
0.21
|
100
|
2.13*
|
11.459*
|
0.13
|
4.11α
|
0.014
|
+ Significant at 90% confidence Level * 95% confidence level, ∞ 99% confidence level, α 99.99% confidence level
Annual
|
Winter
|
Pre-Monsoon
|
Monsoon
|
Post-Monsoon
|
Table 2
Mean (M), standard deviation (SD), coefficient of variation (CV), kurtosis (CK), skewness (CS), Mann-Kendall test (Z), Sen’s slope estimator (Q), and regression (R) of monthly, seasonal, and annual rainfall over ecologically sensitive areas along the Western Ghats (1981–2020)
Excess
|
Deficit
|
Excess
|
Deficit
|
Excess
|
Deficit
|
Excess
|
Deficit
|
Excess
|
Deficit
|
1981
|
1985
|
1984
|
1982
|
1999
|
1983
|
1981
|
1985
|
1987
|
1983
|
2006
|
1986
|
1985
|
1983
|
2004
|
1991
|
2007
|
1986
|
1993
|
1986
|
2007
|
1987
|
1994
|
1987
|
2006
|
1996
|
2019
|
1987
|
2006
|
1988
|
2010
|
2000
|
2000
|
1988
|
2008
|
1997
|
2020
|
1999
|
2010
|
1991
|
2013
|
2003
|
2003
|
1989
|
-
|
2019
|
-
|
2000
|
2019
|
2016
|
2019
|
-
|
2011
|
1991
|
-
|
-
|
-
|
2002
|
-
|
-
|
2020
|
-
|
2013
|
1997
|
-
|
-
|
-
|
2003
|
-
|
-
|
-
|
-
|
-
|
2007
|
-
|
-
|
-
|
2015
|
-
|
-
|
-
|
-
|
-
|
2009
|
-
|
-
|
-
|
-
|
-
|
-
|
Sr. No.
|
Very Strong El-Nino
|
Strong El-Nino
|
Moderate El-Nino
|
Weak El-Nino
|
Neutral
|
Weak La-Nina
|
Moderate La-Nina
|
Strong La-Nina
|
|
Categories
|
Table 3
Excess and deficient rainfall years over ecologically sensitive areas along the Western Ghats (1981–2020)
1
|
1982
|
1987
|
1986
|
1977
|
1978
|
1984
|
1989
|
1988
|
2
|
1983
|
1997
|
1991
|
1979
|
1980
|
1985
|
2011
|
1998
|
3
|
2015
|
-
|
2004
|
2005
|
1981
|
1993
|
-
|
2007
|
4
|
2016
|
-
|
-
|
2009
|
1990
|
1995
|
-
|
2010
|
5
|
-
|
-
|
-
|
-
|
1992
|
1999
|
-
|
-
|
6
|
-
|
-
|
-
|
-
|
1994
|
2000
|
-
|
-
|
7
|
-
|
-
|
-
|
-
|
1996
|
2001
|
-
|
-
|
8
|
-
|
-
|
-
|
-
|
2002
|
2008
|
-
|
-
|
9
|
-
|
-
|
-
|
-
|
2003
|
-
|
-
|
-
|
10
|
-
|
-
|
-
|
-
|
2006
|
-
|
-
|
-
|
11
|
-
|
-
|
-
|
-
|
2012
|
-
|
-
|
-
|
12
|
-
|
-
|
-
|
-
|
2013
|
-
|
-
|
-
|
13
|
-
|
-
|
-
|
-
|
2014
|
-
|
-
|
-
|
Table 4 List of El-Nino and La-Nina years
Source: Mann and Gupta 2022
Stations that lie between 14o N and 16o N (Honavar has the highest increase in rainfall, i.e., 31.87 mm/year), Dandeli, Sagara, and the area around 19o N latitude (Alibag, Bhira, Dahanu, and Santacruz) exhibited a significant increasing trend in both rainfall and temperature (Table 5). Further, Z and Q values in Table 5 show different trend levels in both rainfall and temperature. The results of our study match the results of a study conducted by Francis and Gadgil (2006). Also, rainfall over the study area increased over the time period (1981–2020). When the temperature was averaged over different stations in the study area, it showed a significant increasing trend at all the stations from 1981 to 2020 (Table 5). Trends in temperature, as shown in Fig. 4, reflect an increase in both maximum and minimum temperatures. The increase in minimum temperature extremes in the study area is larger than the increase in maximum temperature extremes. Also, the rate of change of the trends is higher for minimum temperature. The above results are in agreement with previous studies that used monthly data. These studies are Khothawale and Kumar (2005); Ravindranath et al. (2006); and Dhorde et al. (2009). The recent rise in temperature in the study area is due to an increase in anthropogenic activities like mining and quarrying, LULC changes, industrialization, and urbanization. As a result of the above-discussed factors, the study area often experiences disasters and floods like those in Mumbai (2005), Kerala (2018), the annual floods in Mumbai and Pune (2019), and other areas.
Seasonal
Seasonal temporal trends of rainfall over the study area are shown in Fig. 5. The monsoon season receives nearly 80% of annual rainfall; the rest of the season receives the remaining 20%. Winter is less variable season (SD = 13.82) because most of the rainfall in the study area is experienced in the monsoon months (SD = 317.06) (Table 2). On the contrary, CV values are high during winter (102.16%) and low in the monsoon season (18.21%). Furthermore, results obtained from the non-parametric MK test show a prevalence of a positive trend in all seasons, but a significant (95% confidence level) increase was observed only in the post-monsoon season (1.98 mm/year). However, the slope (Q value) is high in the case of monsoon season, which is about 0.019o C/year (Table 2). It is clearly seen in Fig. 5 that rainfall increases in all seasons except winters. Also, the 5-year moving average follows the cyclic pattern in the monsoon and post-monsoon seasons. Station-wise Z and Q values of both rainfall and temperature with their respective significant levels are shown in Table 5. In case of rainfall, the winter and post-monsoon (except Kovalam) seasons show no trend. However, in the pre-monsoon season, significant positive trends were observed in Bandipur, Kannur, Kozhikode, Madikeri, Mangaluru, and Thrissur, and in the monsoon over Alibag, Bhira, Dahanu, Dandeli, Honavar, Sagara, and Santacruz. Temperature showed a significant positive trend in the pre-monsoon, monsoon, and post-monsoon seasons almost everywhere in the study area. Whereas, a significant positive trend was observed over Kannur, Kozhikode, and Kumarakom only in case of winter season.
Monthly
The mean monthly rainfall and temperature over ecologically sensitive areas over the WG is presented in Table 2. Analysis of monthly rainfall reveals that it is highly erratic all over the study area. Only three months (June, July, and August) accounted for more than 67.25% of total rainfall in the study area, whereas the lowest rainfall was received in the month of January (0.26%). Both SD and CV values found high in monsoon months reveal that rainfall is highly variable during these months. However, the Q (Sen Slope) statistic shows that rainfall is decreasing in the months of January and July while it is increasing in the rest of the months. A significant positive trend was observed only in September (3.87 mm/year at 95% confidence level); the rest of the months show no trend. Furthermore, the Z statistic demonstrates that temperatures increase in all the months, but a significant positive increase is noticed in nine months except January, May, and December. The highest increase was observed in the month of August (0.0210 C/year), which is statistically significant at 99.99% confidence level. The month-wise details of both Z and Q values are shown in Table 2.
4.2 Spatial Analysis
Annual
The spatial distribution of average annual rainfall, variability of rainfall, percent change, and rainfall trends over ecologically sensitive areas along the WG during 1981–2020 is shown in Fig. 6. The average annual rainfall has shown an increasing trend from south to north (up to Goa); afterwards, it is decreasing up to Dahanu (except in areas around Mumbai and Alibag). Rainfall occurrence varies across the entire study area, with rainfall less than 2000 mm in the northern and southern parts and more than 2275 mm in the central part of WG. This decline in rainfall from south to north is mainly because the elevation and onset or withdrawal span of monsoon winds are higher in the southern part of WG (Gadgil 2011). The lowest rainfall has occurred over Belagavi station (1514.38), whereas the highest has occurred over Manjeri station (3474.79) during the 40-year study period (Table 1). The spatial pattern of CV as shown in Fig. 6 depicts that the entire study area lies in the moderate variability zone (Hare 2003). Furthermore, CV values increase from south to north, with the lowest at Udupi (17.27) and the highest at Dahanu (26.91). It is a well-established fact that annual variability in rainfall is less in areas of heavy rainfall. Areas situated in the northern and central parts have observed the maximum percent change in average annual rainfall (Fig. 6). Honavar station has shown the highest percent increase (79%) in average annual rainfall during the study period (1981–2020). Furthermore, trends in annual rainfall calculated using the MK test and Sen’s slope estimator show that all the stations except Ratnagiri experienced a positive trend in average annual rainfall. Among all the 27 stations, nine stations, namely Alibag, Bhira, Dahanu, Dandeli, Honavar, Manjeri, Sagara, Santacruz, and Thrissur, show statistically significant positive trends. The magnitude of the annual increase in rainfall is greatest over the Honavar station (31.87 mm/year at a 99.99% significant trend). Majority of stations located in the northern and central parts of the study area registered a positive trend, i.e., a significant trend, which coincides with stations having heavy rainfall events. It is a source of concern for the entire vulnerable region of WG, making the area vulnerable to flooding. The mean maximum, minimum, and average temperature over ecologically sensitive areas along the WG for the period 1981–2020 is shown in Fig. 7. The spatial pattern of average annual temperature shows that the increase in temperature was highest in the northern part and along the coastal belt of Karnataka and Kerala. Also, the increase in temperature was high around Sirsi, Honavar, and Sagara. The distribution pattern, however, differs in case of minimum temperature. It is increasing around Harnai, Bhira, Ratnagiri, and in the southern part, around coastal Kerala. In the case of maximum temperature, an increase in temperature was noticed in small pockets around Mumbai, including Sirsi, Honavar, Kumarakom, and Painavu.
Seasonal
The seasonal distribution of rainfall, percent change and rainfall trends over the study area are shown in Fig. 8. During the winter season, the entire study area receives an average rainfall of about 13.89 mm. The southern part of Kerala receives the highest rainfall, which decreases towards the north. During the pre-monsoon season, rainfall patterns are almost identical except for a marginal increase in area under each category. The northern part of WG in the pre-monsoon season comes under the category of less than 100 mm of rainfall, whereas the southern part comes under 300 to 400 mm of rainfall.
The distribution of rainfall in case of monsoon season is highest in the central part of the study area, and a consistent decreasing trend was observed in the northern and southern parts. The amount of rainfall has ranged from > 2500 mm in the central parts to less than 1000 mm in the southern parts of the study area. Nearly 80% of annual rainfall is received in the monsoon season; hence, the rainfall distribution during the monsoon months is almost identical to the average annual rainfall. Due to the difference in elevation between the northern and southern parts of WG, rainfall is reduced from south to north. Furthermore, in the post-monsoon season, a proper decreasing trend in rainfall is noticed from south to north. In contrast, the spatial pattern of CV in different seasons has shown a reverse pattern in the seasonal distribution of rainfall. The variation in CV values across the study area is less during monsoon season because rainfall is abundant during these months. However, the values of CV in other seasons, i.e., winter, pre-monsoon, and post-monsoon, are high due to the occurrence of scanty and inconsistent rainfall. Spatial distribution maps of percent change show that during the winter season, in the central part of the study area, only three stations (Belagavi, Dandeli, and Sirsi) have observed a positive change of more than 20%. The northern (Dahanu) and parts of coastal Karnataka (Mangaluru, Madikeri, and Udupi) and Kerala (Kumarakom) have observed a positive change of more than 30% for the period 1981–2020 in the pre-monsoon season. However, in monsoon months, a negative change in rainfall is observed in the southern parts of the study area (Fig. 8). No station in the post-monsoon season shows a negative change in rainfall over the entire study period. Seasonal trends in rainfall are computed using the MK test and Sen’s slope estimator. During the winter season, majority of stations in the study area have shown a non-significant positive trend except Vengurla, Ratnagiri, Honavar, Harnai, Belagavi, Bhira, Dahanu, and Alibag. A significant increase in rainfall is noticed in the southern part of WG (Bandipur, Kannur, Kozhikode, Madikeri, Mangaluru, and Thrissur) during the pre-monsoon season. The magnitude of the increase in rainfall has varied from 3.88 mm/year (Kozhikode) to 2.7 mm/year (Mangaluru).
Stations
|
Rainfall
|
Average Temperature
|
|
Winter
|
Pre-Monsoon
|
Monsoon
|
Post-Monsoon
|
Annual
|
Winter
|
Pre-Monsoon
|
Monsoon
|
Post-Monsoon
|
Annual
|
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Z
|
Q
|
Alibag
|
-1.01
|
-0.01
|
-0.44
|
-0.05
|
2.18*
|
18.11*
|
1.36
|
1.52
|
2.23*
|
17.86*
|
0.83
|
0.008
|
3.16∞
|
0.023∞
|
2.45*
|
0.013*
|
2.24*
|
0.0189*
|
3.87α
|
0.015α
|
Bandipur
|
1.28
|
0.05
|
2.46*
|
3.16*
|
-0.9
|
-4.48
|
0.62
|
1.01
|
0.31
|
2.21
|
1.13
|
0.012
|
-0.28
|
-0.003
|
4.99α
|
0.026α
|
2.14*
|
0.0197*
|
2.56*
|
0.015*
|
Belagavi
|
-0.35
|
0.00
|
0.62
|
0.41
|
0.29
|
1.02
|
0.27
|
0.23
|
0.57
|
2.93
|
0.47
|
0.004
|
2.75∞
|
0.021∞
|
3.48α
|
0.018α
|
1.06
|
0.0123
|
3.78α
|
0.014α
|
Bhira
|
-0.78
|
-0.03
|
-0.8
|
-0.17
|
2.44*
|
12.54*
|
1.48
|
1.38
|
2.44*
|
13.72*
|
1.57
|
0.009
|
3.10∞
|
0.022∞
|
3.39α
|
0.016α
|
2.75∞
|
0.0261∞
|
4.04α
|
0.018α
|
Dahanu
|
-0.72
|
-0.02
|
1.2
|
0.16
|
2.97∞
|
18.18∞
|
0.92
|
0.55
|
3.27∞
|
19.60∞
|
0.03
|
0.001
|
2.88∞
|
0.016∞
|
2.83∞
|
0.014∞
|
1.4
|
0.0113
|
2.89∞
|
0.012∞
|
Dandeli
|
0.08
|
0.00
|
1.11
|
0.83
|
3.2∞
|
19.95∞
|
0.55
|
0.77
|
3.48α
|
21.17α
|
0.79
|
0.006
|
2.61∞
|
0.017∞
|
3.74α
|
0.016α
|
2.11*
|
0.0178*
|
3.46α
|
0.013α
|
Harnai
|
-1.36
|
-0.04
|
-0.41
|
-0.11
|
0.69
|
4.06
|
1.13
|
1.50
|
0.8
|
5.92
|
1.33
|
0.012
|
3.10∞
|
0.022∞
|
3.37α
|
0.016α
|
2.48*
|
0.0199
|
4.37α
|
0.017α
|
Honavar
|
-0.05
|
0.00
|
0.99
|
1.01
|
3.95α
|
31.50α
|
0.76
|
1.09
|
3.9α
|
31.87α
|
0.34
|
0.003
|
2.80∞
|
0.017∞
|
3.52α
|
0.014α
|
2.04*
|
0.0147*
|
3.67α
|
0.013α
|
Kannur
|
0.62
|
0.04
|
2.2*
|
3.65*
|
0.08
|
0.72
|
1.6
|
2.30
|
1.2
|
10.56
|
1.97*
|
0.014*
|
0.33
|
0.001
|
4.72α
|
0.019α
|
2.88∞
|
0.0176∞
|
3.39α
|
0.013α
|
Kovalam
|
0.55
|
0.12
|
0.43
|
0.63
|
-0.64
|
-1.70
|
1.76
|
4.90
|
0.66
|
3.28
|
0.54
|
0.006
|
2.19*
|
0.013*
|
3.78α
|
0.021α
|
1.61
|
0.0095
|
3.18∞
|
0.014∞
|
Kozhikode
|
0.94
|
0.07
|
2.18*
|
3.78*
|
0.15
|
0.74
|
1.99*
|
3.81*
|
1.2
|
8.44
|
2.60∞
|
0.013*
|
1.19
|
0.007
|
4.45α
|
0.019α
|
3.20∞
|
0.0181∞
|
3.85α
|
0.015α
|
Kumarakom
|
0.72
|
0.24
|
1.55
|
3.51
|
0.87
|
5.80
|
1.6
|
4.73
|
1.34
|
8.90
|
2.33*
|
0.013*
|
2.03*
|
0.013*
|
3.86α
|
0.020α
|
3.19∞
|
0.0171∞
|
4.09α
|
0.017α
|
Madikeri
|
0.48
|
0.02
|
2.44*
|
3.13*
|
-0.2
|
-0.95
|
1.06
|
1.45
|
0.85
|
8.91
|
1.61
|
0.012
|
-0.03
|
0.000
|
4.85α
|
0.020α
|
3.10∞
|
0.0183∞
|
2.87∞
|
0.014∞
|
Manjeri
|
1.14
|
0.14
|
1.85
|
2.50
|
-0.08
|
-0.32
|
0.97
|
2.44
|
1.71
|
16.84
|
1.81
|
0.013
|
0.22
|
0.001
|
4.80α
|
0.021α
|
2.26*
|
0.0143*
|
2.88α
|
0.015α
|
Mangaluru
|
0.42
|
0.01
|
1.97*
|
2.70*
|
1.11
|
9.96
|
0.8
|
1.36
|
1.15
|
7.01
|
1.77
|
0.010
|
1.21
|
0.008
|
4.44α
|
0.019α
|
3.38α
|
0.0163α
|
3.81α
|
0.014α
|
Painavu
|
0.7
|
0.18
|
1.32
|
1.93
|
0.29
|
1.36
|
1.69
|
4.49
|
1.01
|
5.03
|
0.86
|
0.008
|
0.15
|
0.002
|
4.44α
|
0.023α
|
2.85∞
|
0.0161∞
|
2.78∞
|
0.014∞
|
Panjim
|
0.79
|
0.00
|
0.41
|
0.22
|
1.57
|
8.63
|
0.45
|
0.41
|
1.62
|
8.06
|
1.27
|
0.008
|
3.04∞
|
0.018∞
|
3.83α
|
0.017α
|
2.09*
|
0.0159*
|
4.85α
|
0.016α
|
Ponmudi
|
0.56
|
0.14
|
0.87
|
1.88
|
-0.38
|
-1.39
|
1.57
|
5.28
|
0.59
|
3.46
|
0.13
|
0.001
|
0.7
|
0.007
|
4.45α
|
0.020α
|
1.91
|
0.0109
|
2.63∞
|
0.014∞
|
Ratnagiri
|
-1.29
|
-0.02
|
0.1
|
0.04
|
-0.9
|
-6.06
|
0.87
|
0.97
|
-0.45
|
-3.49
|
1.57
|
0.012
|
2.98∞
|
0.021∞
|
3.44α
|
0.017α
|
2.74∞
|
0.0217∞
|
4.43α
|
0.017α
|
Sagara
|
0.06
|
0.00
|
0.87
|
0.95
|
1.97*
|
12.17*
|
0.31
|
0.42
|
2.2*
|
12.62*
|
1.55
|
0.015
|
2.20*
|
0.017*
|
4.51α
|
0.020α
|
3.05∞
|
0.0187∞
|
3.73α
|
0.016α
|
Santacruz
|
0.16
|
0.00
|
-0.24
|
-0.06
|
2.41*
|
19.55*
|
1.04
|
1.10
|
2.57*
|
19.74*
|
0.45
|
0.005
|
2.92∞
|
0.020∞
|
2.02*
|
0.012*
|
1.88
|
0.0160
|
3.23∞
|
0.013∞
|
Sirsi
|
0.19
|
0.00
|
0.76
|
0.66
|
0.99
|
5.12
|
0.13
|
0.11
|
1.29
|
5.21
|
0.63
|
0.007
|
2.20*
|
0.015*
|
4.14α
|
0.017α
|
2.13*
|
0.0167*
|
3.04∞
|
0.014∞
|
Thekkady
|
0.47
|
0.13
|
0.85
|
1.31
|
-0.2
|
-1.30
|
1.64
|
4.43
|
0.29
|
1.66
|
0.62
|
0.009
|
0.49
|
0.004
|
4.67α
|
0.022α
|
2.81∞
|
0.0147∞
|
2.84∞
|
0.015∞
|
Thrissur
|
0.98
|
0.22
|
1.97*
|
2.76*
|
0.66
|
3.87
|
1.46
|
3.04
|
1.95
|
11.69
|
1.81
|
0.015
|
0.63
|
0.005
|
4.60α
|
0.020α
|
1.95
|
0.0142
|
2.90∞
|
0.014∞
|
Udupi
|
0.18
|
0.00
|
1.29
|
1.69
|
0.8
|
6.69
|
0.55
|
0.72
|
1.29
|
9.69
|
1.74
|
0.015
|
2.03*
|
0.012*
|
4.64α
|
0.020α
|
3.19∞
|
0.0173∞
|
3.92α
|
0.016α
|
Vengurla
|
-0.66
|
0.00
|
0.59
|
0.49
|
0.5
|
2.87
|
0.83
|
0.74
|
0.66
|
5.16
|
1.31
|
0.009
|
3.62α
|
0.021α
|
3.59α
|
0.017α
|
1.69
|
0.0131
|
4.78α
|
0.016α
|
Wayanad
|
1.12
|
0.09
|
1.92
|
2.83
|
-0.5
|
-2.49
|
0.92
|
1.67
|
0.43
|
2.37
|
1.41
|
0.014
|
-0.08
|
-0.001
|
5.07α
|
0.022α
|
2.35*
|
0.0182*
|
2.81∞
|
0.016∞
|
* 95% confidence level, ∞ 99% confidence level, α 99.99% confidence level |
Table 5 Station-wise Z (Mann-Kendall) and Q values (Sen’s slope) of seasonal and annual rainfall and average temperature over ecologically sensitive areas along the Western Ghats (1981–2020)
During the monsoon season, only around 14oN to 16oN and around 19oN, i.e. the area around the stations of Alibag, Bhira, Dahanu, Dandeli, Honavar, Sagara, and Santacruz, observed a significant increase in rainfall. The rate of increase in rainfall has varied from 31.50 mm/year (Honavar station) to 12.17 mm/year (Sagara station) (Table 5). During the post-monsoon season, a significant increase in rainfall was observed only over Kozhikode, Painavu, and Kovalam (Fig. 8).
Monthly
The spatial distribution of mean monthly rainfall during 1981–2020 is shown in Fig. 9. In the first quarter, i.e., in the months of January, February, and March, rainfall in the northern part of the study area is low, while it is high in the southern part and the amount of rainfall is almost negligible. Except for a slight increase in rainfall in May, the distribution of rainfall in April and May was nearly identical. In the next four months, rainfall patterns change suddenly, both in terms of intensity and occurrence. During the entire span of these months, the occurrence of rainfall has been increasing in the northern part until August, and then suddenly it decreases in the month of September. Again, the pattern of rainfall is shifting southward (from October to December), i.e., high rainfall in the southern part and low in the northern part. The CV values over the entire study area are highest in the month of March (149.32) and lowest in the month of June (22.66) (Table 2). The spatial pattern of CV, on the other hand, shows a reverse pattern for mean monthly rainfall.
The distribution of the monthly maximum temperature is shown in Fig. 10. We take the maximum temperatures of January, June, and October because the rate of increase in temperature is highest in these three months. It is clearly seen in the maps of January that the temperature shows an increasing trend all over the study area in the category of above 32o C. However, the highest increase was observed in the northern part of WG. Most likely, it is the reason behind the occurrence of the heaviest rainfall events in the northern and central parts of the study area. In June, the increase in temperature is high around the stations of Alibag, Santacruz, and Dahanu. Again, in the month of October, the degree and distribution of temperature increase are the same as in June, i.e., the highest increase in temperature around Mumbai.
4.3 LULC transformation
The LULC maps of WG are dominated by open, built-up, agricultural, and forest land (Fig. 11). Throughout the study period from 1991 to 2021, the built-up area has steadily increased, but the area under forest has decreased. Furthermore, between 1991 and 2021, the area under open land has increased significantly. The increase in the area under the category of open land is at the expense of agricultural and forest land. Similarly, the number of water bodies or wetlands has decreased between 1991 and 2021. Table 6 shows the area in square kilometres as well as the proportion of area under each category from 1991 to 2021. This change is mainly linked with the growth of urbanization, industrialization, or the construction of roads and highways, as well as the ongoing growth in areas under the built-up category (around the Mumbai metropolitan region and coastal towns of Kerala) and mining operations in the vicinity of Goa. Calculations of the LULC classes' decadal net change and total net change for the time periods 1991–2001, 2001–2011, 2011–2021, and 1991–2021 were made (Table 7). Positive values indicate gain, whereas negative ones indicate loss. Forest (loss) observed the highest net change between 1991 and 2001, followed by open land (gain), built-up land (gain), agricultural land (gain), and waterbody (loss). However, the values are inverted in the case of the highest net change from 2001 to 2011 and from 2011 to 2021. In the next two decades, it was found that open land (gain), forest cover (loss), built-up land (gain), agricultural land (loss), and waterbody (gain) experienced the highest net changes. The trend of forest degradation has increased significantly during 2001–2011, as indicated by the net change in forest from 1991–2001 to 2001–2011, which ranges from − 1.75 to -9.25. Open land showed the highest total net change, followed by forest cover, built-up land, agricultural land, and Waterbody. However, it is clearly seen in Fig. 12 that the highest percent change is observed in the open land (16.45%) category, followed by forest cover (-15.79%), built-up land (10.23%), agriculture land (-9.11%), and waterbody (-1.78%). The proportion of area under each category is depicted in Fig. 13. It is observed that forest cover is still the dominant class in the LULC classification, but the rate of degradation is very fast. It is a major concern for the ecological balance of the study area. During the period from 1991 to 2021, the built-up area increased primarily at the expense of agricultural and forest land. There are examples of waste or open land and forest (mostly deciduous) being turned into construction sites, particularly in the coastal towns of Kerala and the vicinity of Thane and Pune.
Plantations and related built-up cover the vast majority of land in Kerala. Depending on the built-up's relative importance, a demarcation between plantations and other classes should be made. If plantations are the dominant class in the region, plantations are still taken into account even if there is a built-up understory, and vice versa. A growing built-up dominance in the area is linked to the conversion of plantations to built-up land. Furthermore, from 2001 to 2021, there has been an increase in the proportion of built-up area around the stations of Udupi, Mangaluru, and Honavar (Karnataka), Alibag, Bhira, and Dahanu (Maharashtra), as well as the area around the regions of Dharwad and Shimoga (Karnataka). In Maharashtra, Kerala and Karnataka, between 1991 and 2021, the area under the built-up and open land categories grew mostly at the expense of agriculture and forest land. In low-rainfall regions of Karnataka and Tamil Nadu, the conversion of cropland to plantations was prevalent. The land becomes fallow because people are unable to harvest good crops. Such an uncultivated area gradually becomes a plantation. In comparison to other forest types, the Northern WG's deciduous broadleaf forests are facing a greater threat of degradation. In contrast to Southern WG, the transformation of forest into open land or cropland is more prevalent in Northern WG. There are reported instances of evergreen forest being converted to open land throughout the study area, and the majority of the forest that is changed to open or agricultural land in these locations is primarily deciduous forest. The area around the regions of Dharwad and Shimoga (Karnataka). In Maharashtra, Kerala and Karnataka, between 1991 and 2021, the area under the built-up and open land categories grew mostly at the expense of agriculture and forest land. In low-rainfall regions of Karnataka and Tamil Nadu, the conversion of cropland to plantations was prevalent. The land becomes fallow because people are unable to harvest good crops. Such an uncultivated area gradually becomes a plantation. In comparison to other forest types, the Northern WG's deciduous broadleaf forests are facing a greater threat of degradation. In contrast to Southern WG, the transformation of forest into open land or cropland is more prevalent in Northern WG. There are reported instances of evergreen forest being converted to open land throughout the study area, and the majority of the forest that is changed to open or agricultural land in these locations is primarily deciduous forest.
4.4 Association of rainfall and temperature with LULC
Changes in the patterns and distribution of temperature, rainfall, and heavy rainfall events can be attributed to either natural or anthropogenic factors, or a mixture of the two. Climate change, global warming, and environmental degradation have all recently become more prevalent, which is blamed for the natural change. Human-induced LULC change and unchecked urbanisation may have an impact on surface heat influxes, which in turn effect regional atmospheric circulation. Despite being static, topographic factors like elevation, slope, and aspect also contribute to and affect the regional distribution of trends in both rainfall and temperature. Also, various other meteorological parameters have an impact on how rain behaves globally. The influence of LULC change is noticed in the changing climatic trends of rainfall, temperature, and heavy rainfall events in the study area.
Figure 6 Spatial distribution of average annual rainfall, coefficient of variation, percent change, and rainfall trends in rainfall over ecologically sensitive areas along the Western Ghats (1981–2020)
Figure 7 Spatial distribution of average annual, minimum, and maximum temperature over ecologically sensitive areas along the Western Ghats (1981–2020)
Figure 8 Spatial distribution of seasonal rainfall, percent change, and rainfall trends over ecologically sensitive areas along the Western Ghats (1981–2020)
Figure 9 Spatial distribution of monthly rainfall over ecologically sensitive areas along the Western Ghats (1981–2020)
Figure 10 Spatial distribution of maximum temperature in the months of January, June, and October over ecologically sensitive areas along the Western Ghats (1981–2020)
Figure 11 Land use and Land cover transformation map of ecologically sensitive areas along the Western Ghats (1991–2021)
Year
|
1991
|
2001
|
2011
|
2021
|
Land use Land cover classes
|
Area (Km2)
|
Area (%)
|
Area (Km2)
|
Area (%)
|
Area (Km2)
|
Area (%)
|
Area (Km2)
|
Area (%)
|
Waterbody
|
3837.40
|
4.15
|
3602.72
|
3.89
|
2369.42
|
2.56
|
2193.57
|
2.37
|
Agriculture Land
|
15484.57
|
16.73
|
15808.52
|
17.08
|
10218.15
|
11.04
|
7052.75
|
7.62
|
Forest Cover
|
69574.14
|
75.17
|
67954.42
|
73.42
|
59393.03
|
64.17
|
54959.59
|
59.38
|
Built-Up Land
|
1906.65
|
2.06
|
2628.58
|
2.84
|
8950.14
|
9.67
|
11375.10
|
12.29
|
Open Land
|
1752.97
|
1.89
|
2561.49
|
2.77
|
11624.99
|
12.56
|
16974.72
|
18.34
|
Total
|
92555.73
|
|
92555.73
|
|
92555.73
|
|
92555.73
|
|
Table 6 Area (km2) and the ratio of Land Use/Land Cover class of different years (1991, 2001, 2011 and 2021) over ecologically sensitive areas along the Western Ghats
Figure 13 LULC relationship between Classes over ecologically sensitive areas along the Western Ghats (1991–2021)
Table 7 Decadal change in land-use and land-cover area over ecologically sensitive areas along the Western Ghats (1991-2021)
Land-use and Land-cover classes
|
1991
|
2001
|
2011
|
2021
|
Net Change (1991–2001)
|
Net Change (2001–2011)
|
Net Change (2011–2021)
|
Total Net Change (1991–2021)
|
Waterbody
|
4.15
|
3.89
|
2.56
|
2.37
|
-0.25
|
-1.33
|
-0.19
|
-1.78
|
Agriculture land
|
16.73
|
17.08
|
11.04
|
7.62
|
0.35
|
-6.04
|
-3.42
|
-9.11
|
Forest cover
|
75.17
|
73.42
|
64.17
|
59.38
|
-1.75
|
-9.25
|
-4.79
|
-15.79
|
Built-up land
|
2.06
|
2.84
|
9.67
|
12.29
|
0.78
|
6.83
|
2.62
|
10.23
|
Open land
|
1.89
|
2.77
|
12.56
|
18.34
|
0.87
|
9.79
|
5.78
|
16.45
|
Figures are in percentage |
Table 8 Stations experiencing maximum change in rainfall, temperature and heavy rainfall events in association with LULC change over ecologically sensitive areas along the Western Ghats (1981-2020)
Latitudes experiences maximum change in LULC
|
Heavy Rainfall Events
|
Rainfall
|
Temperature
|
Significant Increase
|
Minimum
|
Maximum
|
Average
|
18o to 19o N
|
Dahanu
|
Dahanu
|
Dahanu
|
Dahanu
|
Dahanu
|
Santacruz
|
Santacruz
|
Santacruz
|
Santacruz
|
Santacruz
|
Alibag
|
Alibag
|
Alibag
|
Alibag
|
Alibag
|
Bhira
|
Bhira
|
Bhira
|
Bhira
|
Bhira
|
Harnai
|
-
|
Harnai
|
-
|
-
|
14o to 16o N
|
Vengurla
|
Vengurla
|
Ratnagiri
|
Sirsi
|
Sirsi
|
Panjim
|
Panjim
|
-
|
Honavar
|
Honavar
|
Dandeli
|
Dandeli
|
-
|
-
|
Sagara
|
Honavar
|
Honavar
|
-
|
-
|
-
|
Udupi
|
Sagara
|
-
|
-
|
-
|
9o to 12o N
|
Mangaluru
|
Manjeri
|
Kannur
|
Painavu
|
Kannur
|
Madikeri
|
Thrissur
|
Manjeri
|
Kumarakom
|
Kozhikode
|
Kumarakom
|
-
|
Kumarakom
|
-
|
Manjeri
|
Kovalam
|
-
|
Kovalam
|
-
|
Thrissur
|
-
|
-
|
-
|
-
|
Kumarakom
|
-
|
-
|
-
|
-
|
Kovalam
|