Annual and seasonal trends in Actual Evaporation over different meteorological sub-divisions in India using GLEAM data

DOI: https://doi.org/10.21203/rs.3.rs-1465198/v1

Abstract

The present study analyzed the actual evaporation (Ea) and its components [transpiration (Et), bare soil evaporation (Eb), interception loss (Ei), and open water evaporation (Eo)] data to study the long-term (1980-2018) trends over different meteorological sub-divisions in India. Quantitatively, all India annual Ea is 573 mm (σ= 28.8 mm), where Et (µ=455.9 mm; σ= 30.4 mm) plays a major role compared to other evaporation processes like Eb (µ=55.9 mm; σ= 8.6 mm), Ei (µ=33.5mm; σ= 3.4 mm), and Eo (µ=27.2 mm; σ=0.38 mm). The MK test reveals an increasing trend (1.33mm/y) in annual Ea due to the rising trend in Et (1.91mm/y) and Ei (0.16mm/y) even though there is no significant trend in rainfall or potential evaporation (Ep). The sub-division-wise analysis shows the increasing trend in Ea observed over irrigated regions located in the south, north-west, and foothills of the Himalayas during pre-monsoon and monsoon season. The correlation analysis observed a complex relationship between Ea and climatic factors (rainfall (RF), soil moisture (SM), surface temperature (T), relative humidity (RH), surface shortwave radiation (SSR), and wind speed (WS)) during monsoon season such that the water-limited areas have a positive correlation with SM, RH and RF, and negative correlation with WS, T, and SSR, whereas, in energy-limited areas (east India), the Ea showed a positive correlation with SSR & T and negative correlation with RF. The main climatic drivers for the increasing trend of Ea are SM & rainfall over dry regions and SSR &T over densely vegetated regions in India.

1. Introduction

Spatio-temporal variability of actual evaporation (Ea) plays a tremendous role in the global energy budget by altering Bowen’s ratio, hydrological cycle, water availability, near-surface temperature, humidity, cloud formation, vegetation health, and drought assessment (Konapala et al., 2020; Rehana and Monish, 2020, 2021). In the case of India, some of the researchers emphasized that the terrestrial evaporation anomalies play a significant role in Indian Summer Monsoon Rainfall (ISMR). They also identified that the regional changes in vegetation, soil moisture, land use land cover changes, and modern irrigation methods are the main driving parameters for this terrestrial evaporation changes over India (Niyogi et al., 2010; Paul et al., 2016; Kantharao and Rakesh, 2018; Budakoti et al. 2021).

Actual Evaporation is the combination of four major components including transpiration (Et), bare soil evaporation (Eb), interception loss (Ei), and open water evaporation (Eo) (Miralles, D. G., et al., 2011). The evaporation process removes the water from the surfaces like soils, wet vegetation, and water bodies whereas the transpiration process removes the water from vegetation through the stomata to the atmosphere (Zotarelli et al., 2010). The major driving meteorological parameters involved in the evapotranspiration process are rainfall, soil moisture, net surface radiation, temperature, humidity, wind speed, vegetation type, leaf area, stomatal conductance, and atmospheric CO2 (Allen et al., 1998; Yadav et al., 2016). The relative importance of these parameters on actual evaporation mainly depends on the geographical conditions, seasons, and large-scale circulation of the atmosphere.

Globally, the evapotranspiration (ET) processes utilize more than 50% of the solar radiation absorbed by the land surface and return more than 60% of the surface precipitation to the atmosphere (Oki and Kanae, 2006; Trenberth et al., 2009; Zeng et al., 2012). There are a lot of uncertainties in the estimation of these land evaporation components at a global scale, some of the researchers were estimated and reported that the global average land ET is ranging between 558 and 650 mm/year (Jung et al., 2010; Zeng et al., 2012). Zhang et al., (2016) studied the multi-decadal trends in global land ET and its components and observed that there was a significant increasing trend over eastern Asia, India, and some other parts of the globe due to the positive trends in transpiration from vegetation and direct evaporation from soil during the period 1981–2012. Zhang et al. 2017, also studied the regional eco-hydrological links and observed that the global actual evaporation is dominated by the direct evaporation from soil surface over dry regions and transpiration over wet regions with high leaf area index. Recently, Zhan et al., 2019 emphasized that the changes in global surface water area due to the impoundment of water by dams were also largely responsible for positive trends in global terrestrial evaporation. Apart from these regional changes, large-scale oscillations (teleconnections) such as El Nino Southern Oscillation (ENSO) and North Atlantic Oscillations (NAO) also play a role in global terrestrial evaporation by altering the regional scale precipitation, temperature, and Humidity (Martens et al, 2018).

Very few researchers have studied the actual evaporation trends and their response to climate change in India. Some of the earlier studies utilized the pan evaporation measurement to study the evaporation trends in India (1961–2000) based on the available data and concluded that the declining trends in India (Chattopadhyay and Hulme (1997); Jaswal et al., 2008; Bandyopadhyay et al., 2009; Padmakumari, Jaswal and Goswami, 2013). There are a lot of uncertainties involved to interpret actual evaporation changes based on the pan evaporation measurements over water-limited and energy-limited areas (Hobbins, Ramírez, and Brown, 2004; Roderick, Hobbins, and Farquhar, 2009). However, the recent studies are utilizing the evapotranspiration estimations based on the satellite-remote sensing observation due to the limited pan evaporation observation points all over India. Goroshi et al. (2017) utilized the actual evaporation product from NTSG (Numerical Terradynamic Simulation Group), the University of Montana, and found that there was a net declining trend in ET over India, while the increasing trend observed over croplands in the central and western part of India during the period 1983–2006. Soni and Syad (2021) estimated the ET values over major river basins in India using the WaterGAP Global Hydrology Model (WGHM) and observed a significant increasing trend over Ganga (2.72mm/year) and Krishna (3.9 mm/year) during the period 1982–2014.

Hence, there is a knowledge gap in quantitative assessment of actual evaporation and its components trends under changing climatic conditions and anthropogenic changes such as rise in irrigated cropland area (Ambika et al. 2016, 2019), surface water area (Zhan et al., 2019), and vegetation (Bhimala et al. 2020) over different meteorological sub-divisions (India Meteorological Department; Kelkar & Sreejith, 2020) in India (Fig. 1) The novelty of the present study is that the actual evaporation data product (Global Land Evaporation Amsterdam Model) utilized in this study is a long-term (1980–2018), consistent, and the satellite observations (soil moisture and vegetation optical depth) data is used for estimation of evaporative stress in the model algorithms (Martens et al., 2017). The study also identifies the relationship between the climatic parameters and actual evaporation over different meteorological sub-divisions in India.

2. Data

2.1 GLEAM (Global Land Evaporation Amsterdam Model) Evaporation data

The long-term (1980–2018) actual evaporation and its components (Et, Eb, Ei, Eo and Ep) data utilized in the present study is adopted from GLEAM (Version3) gridded (0.25°x0.25°) data product (https://www.gleam.eu/ ). The model divides each grid-cell into four fractions based on the MODIS (Moderate Resolution Image Spectroradiometer) land cover type (bare soil, low vegetation, high vegetation, and open water) and calculate the evaporative flux for each fraction based on land cover type and aggregated them to generate a pixel value. Initially, the model calculates the potential evaporation based on the Priestley and Taylor (1972) equation and converts it to actual evaporation based on the evaporative stress factor (0–1) which is the function of vegetation optical depth (VOD) and root zone soil moisture (RZSM). The satellite microwave observations were played a key role in the generation of VOD and RZSM data. The model estimated actual evaporation data has been validated with the in-situ observations globally and results are found satisfactory.

2.2 Climate Data

The high resolution (0.25°x0.25°) daily gridded rainfall data (1980–2018) utilized in the present study is collected from IMD (India Meteorological Department), India (https://imdpune.gov.in/Clim_Pred_LRF_New/Grided_Data_Download.html), Pai et al. (2014). The satellite-based (microwave sensors) soil moisture data product was collected from EAS-CCI (European Space Agency - Climate Change Initiative Soil Moisture; https://www.esa-soilmoisture-cci.org/) (Dorigo et al., 2017; Gruber et al., 2019). The basic climate parameters like temperature (T), Relative Humidity (RH), surface net solar radiation (SSR), and wind speed (WS) data were collected from high resolution (0.1°x0.1°) ERA5 reanalysis data (https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-land-monthly-means?tab=overview). The monthly high resolution (0.05°x0.05°) gridded NDVI data collected from MODIS (Moderate Resolution Imaging Spectroradiometer) (MOD13C2 collection 6), Integrated Climate Data Center (ICDC, icdc.cen.uni-hamburg.de), University of Hamburg, Germany.

3. Methodology

3.1 The Mann-Kendall Test (MK test):

The MK trend test (Gilbert, 1987; Gocic and Trajkovic, 2013) was utilized to compute the monotonic trend (increasing/decreasing) in the actual evaporation and other climatic parameters during the period 1980–2018. The advantage of the MK test is that the collected data need not follow any particular distribution (non-parametric test) and missing values doesn’t harm the performance of the test. The monotonic upward (or downward) trend indicates the increasing trend in the selected data and the calculated trend may or may not be linear. Numerous studies have been utilized this test to calculate the trends in hydrological, climatic parameters (Chaudhuri and Dutta, 2014; Kaur, Diwakar, and Das, 2017; Phuong et al., 2020).

3.2 Standardized Anomaly:

The standardized anomaly (σs) is calculated to identify the excess and dry years from the long-term actual evaporation data, and calculated by dividing the sample anomaly with its standard deviation (σ). The anomaly of the actual evaporation (X) is calculated by subtracting the actual value of the respective year with its long-term average (µ) (Wilks, 2011). The mathematical formula for the standardized anomaly is

$${\sigma }_{s}=\frac{(X-\mu )}{\sigma }$$

Where the standard deviation (σ) is calculated as

$$\sigma =\sqrt{\frac{1}{(N-1)}\sum _{i=1}^{N}{\left({X}_{i}-\mu \right)}^{2}}$$

Here, i represent the corresponding year and the N represents the total number of years.

The magnitude of the anomalies is better represented with standardized anomalies because of the influence of dispersion removed from the sample data. The value of -2 (+ 2) for σs means that the actual evaporation is two standard deviations below (above) normal for that corresponding year.

3.3 Pearson Correlation Coefficient

The correlation coefficient (r) between the actual evaporation (X) and the climatic parameters (Y) is calculated as

Here, i represent the year and the total number of years represented as N (39). The correlation coefficient tells us the strength and direction of the linear relationship between the two parameters. The values are ranging between − 1 and + 1; the positive (negative) value indicates that the climatic parameter has a positive (negative) linear relationship with AET during the study period. It is also noted that Pearson's correlation method depends on the mean of the parameter which means that the coefficient values have more validity for normally distributed data. The correlation significant level (95% or 99%) is calculated based on the p-value, and the p-value calculated from the test statistic t score and the t score is calculated as

4. Results

4.1 Annual trends in all India Evaporation:

Figure 2a shows the climatological annual cycle of the potential evaporation (Ep), actual evaporation (Ea), and its components including transpiration (Et), bare soil evaporation (Eb), interception loss (Ei), and open water evaporation (Eo) over India (Area Average) during the period 1980–2018. The annual cycle shows that there is a huge difference between potential and actual evaporation during the pre-monsoon season compared to the other seasons. The annual cycle also indicates that transpiration is the major component of actual evaporation for most of the period in a year whereas the remaining evaporation components contribute considerably during the summer monsoon season (Fig. 2a). The annual cycle (climatology) of rainfall, soil moisture, and vegetation (MODIS NDVI: 2000–2016) presented along with the actual evaporation to understand the role of climatic factors on an annual cycle of actual evaporation during the study period (Fig. 2b, 2c, and 2d). The time series plot depicts that the maximum amount of actual evaporation was observed on the last week of September (~ 2.5mm/day) whereas the maximum amount of rainfall (9-10mm/day) was observed during the second fortnight of July and the first week of August in an annual scale. The annual cycle of MODIS NDVI (vegetation) shows the maximum (0.595) amount of vegetation observed in September over India. The annual cycle of Ea clearly shows that the terrestrial evaporation is mainly driven by the vegetation and soil moisture dynamics compared to the rainfall over India.

Statistically, all India climatological annual actual evaporation is 573 mm/year with the standard deviation (σ) of 28.8 mm/year from 1980 to 2018. This shows that 51.3% of the total precipitation (1116.5 mm/year) is evaporating back to the atmosphere by different land evaporation ways including transpiration (455.9mm/year), bare soil evaporation (55.9mm/year), interception loss (33.5mm/year), and open water evaporation (27.2mm/year) (Table 1). However, the annual potential evaporation is 966.6mm/year, only, 59.3% of the actual evaporation is taking place due to the lack of surface wetness over many sub-divisions in India.

Table 1

All India mean, standard deviation (σ), and slope of potential and actual evaporation along with its components during pre-monsoon, monsoon, and post-monsoon seasons during the period 1980–2018.

   

Potential Evaporation(mm)

Actual Evaporation(mm)

Transpiration(mm)

Bare soil Evaporation (mm)

Interception loss (mm)

Open water Evaporation (mm)

Annual

Mean(µ)

966.63

573.14

455.90

55.91

33.54

27.28

σ

11.86

28.81

30.43

8.62

3.44

0.38

slope

0.16

1.33

1.91

-0.53

0.16

-0.02

Pre Monsoon (MAM)

Mean(µ)

309.40

104.42

82.51

7.63

5.38

8.83

σ

3.99

13.46

11.77

2.07

0.85

0.11

slope

0.19

0.54

0.47

0.01

0.05

-0.004

Monsoon

(JJAS)

Mean(µ)

359.65

270.27

199.89

37.66

22.87

9.52

σ

11.84

10.62

12.41

5.62

2.24

0.32

slope

-0.27

0.52

0.85

-0.37

0.08

-0.01

Post Monsoon

(OND)

Mean(µ)

182.93

142.44

124.82

8.35

3.84

5.35

σ

3.11

9.72

9.07

2.32

0.77

0.11

slope

0.07

0.30

0.43

-0.11

0.01

-0.002

The inter-annual variability shows that there is an increasing trend (1.3mm/year) in all India actual evaporation even though there are no significant trends in annual rainfall or potential evaporation over India during the period 1980–2018. This increasing trend in actual evaporation is due to the significant increasing trend in transpiration (1.9mm/year) and interception loss (0.16mm/year) even though there is a significant decreasing trend in bare soil evaporation (-0.53mm/year), and open water evaporation (-0.016mm/year). The standardized anomalies (σa) are calculated during the study period and the anomalies are used to monitor the excess and deficit years in Ea during the study period. The analysis observed seven deficit (σs < 1.0) years (1980, 1985, 1989, 1991, 1992, 2002, and 2003) and eight excess (σs > 1.0) years (1998, 2006, 2007, 2008, 2010, 2011, 2013, and 2015) over India during the study period.

4.2 Spatial variability in annual Evaporation:

The spatial maps of potential & actual evaporation describe that the difference between Ep and Ea is less over highly dense-vegetated regions located in the Western Ghats and North-east India, and high over dry regions located in north-west India (Fig. 3a &3b). The sub-division-wise analysis in annual actual evaporation shows that the extremely low (< 250mm) values found over West-Rajasthan (mostly arid & semi-arid region) and Jammu & Kashmir, and low (250-500mm) amount of Ea observed over East-Rajasthan, Saurashtra &Kutch, Gujarat, Madhya Maharashtra, West Madhya Pradesh, Har. Chd. and Delhi (Haryana, Chandigarh and Delhi) sub-divisions. The high (750-1000mm) amount of Ea observed over Tamilnadu & Puducherry, Arunachal Pradesh, Assam & Meghalaya, and very high (> 1000mm) values observed over Kerala, Coastal Karnataka, and NMMT (Nagaland, Mizoram, Manipur, and Tripura) sub-divisions in India (Fig. 3b). However, moderate actual evaporation (500-750mm) values are found over the sub-divisions located in interior parts of south India, foothills of the Himalayas, and East India regions. The spatial maps of evaporation components, however, show that most of the actual evaporation is contributed by the transpiration (Fig. 3c) over most of the sub-divisions, bare soil evaporation played a significant role (~ 10–20% of Ea) in arid and semi-arid regions of India (Fig. 3d), interception loss contributed significantly (~ 10–30% of Ea) over the Western Ghats and North-East India (Fig. 3e), and open water evaporation contributed significantly (~ 10–20% of Ea) over Coastal Andhra Pradesh, Tamilnadu, and Saurashtra & Kutch regions (Fig. 3f).

4.3 Sub-division-wise trends in Annual Evaporation

Trends in annual potential evaporation, actual evaporation, and its components are presented in Fig. 4. The Ep shows a significant increasing trend in a few of the sub-divisions including west-Rajasthan, Saurashtra & Kutch, Madhya Maharashtra, Marathwada, Uttar Pradesh, and Bihar, whereas decreasing trend observed over some of the sub-divisions located in south India including Coastal Andhra Pradesh, Tamil Nadu, Kerala, and Coastal Karnataka (Fig. 4a). The actual evaporation shows a significant increasing trend over the sub-divisions located in foothills of Himalaya (Jammu &Kashmir, Himachal Pradesh, Uttar Pradesh, Bihar, Sikkim, and Gangetic West- Bengal), north-west India (Punjab, Har. Chd. And Delhi, Rajasthan, Saurashtra &Kutch, Gujarat), Western Ghats (Konkan & Goa, Coastal Karnataka, Kerala), and lower parts of South India (Tamil Nadu and South Interior Karnataka) (Fig. 4b). However, the transpiration has shown a significant increasing trend over most of the sub-divisions except Kerala, Himachal Pradesh, Uttarakhand, Chattisgarh, Odisha, Assam, and NMMT (Fig. 4c). In contrary to the transpiration, bare soil evaporation showed a significant decreasing trend over most of the sub-divisions except west-Rajasthan, Saurashtra & Kutch, Konkan & Goa, and Uttarakhand, and the increasing trend observed over western Himalaya region including Jammu & Kashmir and Himachal Pradesh (Fig. 4d). The contribution of interception loss is very minimal in most of the sub-divisions in India, whereas significant increasing trends were observed over a considerable number of sub-divisions located in the foothills of Himalaya, north-west, east, and south India regions (Fig. 4e). In the case of the open water evaporation, the sub-divisions located on the east coast of India contributed more whereas the decreasing trends were observed during the study period (Fig. 4f).

4.4 Seasonal trends in all India Evaporation

Seasonal analysis was carried out for pre-monsoon (March-May), monsoon (June-September), and post-monsoon (October-December) to understand the seasonal Ea trends over India (India as a whole). The seasonal average Ea values are 104mm (σ = 13.4mm), 270mm (σ = 10.6mm), and 142mm (σ = 9.7mm) during pre-monsoon, monsoon, and post-monsoon seasons respectively during the study period. The Mann-Kendell trend analysis shows that there is a significant increasing trend in seasonal Ea during pre-monsoon (0.53mm/year) and monsoon (0.51mm/year), whereas an insignificant (0.29mm/year) trend found during the post-monsoon season for the study period. However, the transpiration from the vegetation showed an increasing trend in all the seasons and contributed major share (pre-monsoon: 79%; monsoon: 74%; Post-Monsoon: 87.6%) during the post-monsoon season. It is also worth noting that the interception loss has shown an increasing trend during the pre-monsoon and monsoon seasons. However, the seasonal potential evaporation shows a significant increasing trend (0.18mm/year) during pre-monsoon, whereas no significant trends were observed during monsoon or post-monsoon season.

4.5 Spatial variability in Seasonal Evaporation

Figure 5 depicts the spatial variability in climatological actual evaporation, soil moisture, and rainfall over different meteorological subdivisions in India during pre-monsoon (March-May), monsoon (June-September), and post-monsoon season for the period 1980–2018. As expected, we have observed a distinct spatial and seasonal variability in actual evaporation, soil moisture, and rainfall climatology over different meteorological sub-divisions in India. However, there was a seasonal variability in evaporation, high amount of evaporation observed over dense vegetation regions (Western Ghats, North East India, and foothills of Himalayas) where the land was covered with woody Savanna, Savanna, and EverGreen Broadleaf forest; the low amount of evaporation observed over arid and semi-arid regions (northwest India) of India (Fig. 5a, 5d, 5g).

During the pre-monsoon season, the actual evaporation values are ranging between 0–1.0, 1.0-1.5, 1.5–2.5, 2.5–3.5 mm/day for 47%, 23.5%, 17.6%, and 11.7% of the sub-divisions (out of 34) respectively (Fig. 5a). The primary reason for this low evaporation rates may be less availability of soil moisture (< 0.3 m3/m3) and vegetation (NDVI < 0.38) due to the very limited (< 1mm/day) amount of rainfall in most of the sub-divisions in India (Fig. 5b, 5c). In the case of the south-west monsoon season, the actual evaporation values are ranging between 1.0-1.5, 1.5–2.5, 2.5–3.5, 3.5–4.5 mm/day for 5.9%, 70.5%, 11.7%, and 11.7% of the sub-divisions respectively (Fig. 5d). The Ea has shown tremendous improvement (> 2mm/day) during the monsoon season due to the high amount of rainfall (> 5mm/day) and soil moisture (> 0.3 m3/m3) over the Western Ghats, central, east, and northeast India (Fig. 5e, 5f). The actual evaporation is ranging between 1.5 to 2.5 mm/day for 56% of the sub-divisions during the post-monsoon season in India (Fig. 5g). It is also observed that the Ea has shown a high amount over the West and East coast of India during the post-monsoon season due to the winter monsoon rainfall (Fig. 5h, 5i).

4.6 Sub-division wise trends in Seasonal Evaporation

Figure 6 depicts the seasonal trends in Ep, Ea and its components during the pre-monsoon, monsoon, and post-monsoon season over different meteorological sub-divisions in India for the period 1980–2018 (Fig. 6). Our analysis observed an increasing trend in potential evaporation over Indo-Gangetic plains, North-West India, and West-Central India regions during the pre-monsoon season (Fig. 6a). In the case of monsoon season, a few numbers (6) of sub-divisions only showed a decreasing trend (Fig. 6b) whereas some of the sub-divisions located in west-central India showed an increasing trend during the post-monsoon season (Fig. 6c). In general, the Ep shows a decreasing trend in South India and an increasing trend in west-central India during pre and post-monsoon season.

In the case of actual evaporation, the MK test reveals that the more number of sub-divisions show an increasing trend during the pre-monsoon season compared to the monsoon and post-monsoon season (Fig. 6d, 6e,6f). The analysis also found that the significant increasing trend (p < 0.05) in pre-monsoon Ea observed over North (Jammu &Kashmir, Himachal, Haryana, and west Uttar Pradesh), North-West (Rajasthan, Gujarat, and west Madhya Pradesh), North-East (Bihar, Sikkim, and Arunachal Pradesh), South India (Coastal Karnataka, Konkan & Goa, South Interior Karnataka, Tamil Nadu, Rayalaseema, Coastal Andhra Pradesh, and Kerala) regions (Fig. 6d). The sub-division wise Sen’s slope values were > 1 mm/season for four sub-divisions in South India [Coastal Karnataka (2.35mm/season); Kerala (1.9 mm/season); South Interior Karnataka (1.4 mm/season)] and two sub-divisions in foot-hills of Himalayas (Har. Chd. & Delhi and SHWB & Sikkim) during the pre-monsoon season.

In the case of monsoon season, an increasing trend in Ea was observed over semi-arid regions (West & east Rajasthan, Rayalaseema, Tamil Nadu, and SI Karnataka) and north & east India (Uttar Pradesh, Bihar, Jharkhand, and Gangetic West Bengal) (Fig. 6e). The Sen’s slope also reveals that the sub-divisions located in arid and semi-arid regions reported the increasing trend of > 1mm/season during the study period. The analysis observed an increasing trend in Ea over foot-hills of the Himalaya and Saurashtra & Kutch regions during the post-monsoon season (Fig. 6f). The study has not observed any decreasing trends in actual evaporation during the study period.

The present study observed a significant difference between potential and actual evaporation trends over India during pre-monsoon, monsoon, and post-monsoon seasons. To understand the reason for the increasing trend in Ea, the seasonal trends are calculated for the major evaporation components including transpiration, bare soil evaporation, interception loss, and open water evaporation during the study period. The transpiration and interception loss from the vegetation have shown a tremendous increasing trend in the sub-divisions where the Ea has shown an increasing trend during the pre-monsoon season (Fig. 6g, 6m). During the monsoon season, even though there is an increasing trend in transpiration (Fig. 6h) and interception loss (Fig. 6n) over most of the sub-divisions in India (where the major crop activities are taking place in plain areas), the Ea has shown an increasing trend (Fig. 6e) in a very few sub-divisions in South India, Northwest India, and East India due to the significant decreasing trend in bare soil evaporation (most of the crop growing areas; Fig. 6k)) and open water evaporation (Coastal Andhra, Telangana, NI Karnataka, Madhya Maharashtra, Marathawada, Gujarat, Saurashtra & Kutch, East Rajasthan, and Uttarakhand; Fig. 6q)). During the post-monsoon season, a very few sub-division which are located in foot-hills of Himalaya shown an increasing trend in Ea (Fig. 6f) due to the increasing trend in transpiration (Uttar Pradesh, Bihar, Sikkim; Fig. 6i)) and open water evaporation (Punjab and Jammu & Kashmir; Fig. 6r)).

4.7 Seasonal Trends in Climatic parameters

Seasonal Trends in climatic parameters comprising of soil moisture (SM), rainfall (RF), surface net solar radiation (SSR), surface temperature (T), wind speed (WS), and relative humidity (RH) were also computed for the study period (1980–2018) to understand the role of climate on actual evaporation anomalies at sub-division scale in India (Fig. 7). Similar to the actual evaporation, significant trends in soil moisture were observed over North-West and south India regions during the pre-monsoon season (Fig. 7a). It was also observed that the pre-monsoon rainfall has shown a significant increasing trend over Western-Ghats (Coastal Karnataka, Konkan & Goa), Kerala, SI Karnataka, and Rayalaseema sub-divisions in India during the period 1980–2018 (Fig. 7d). The decreasing trends in pre-monsoon rainfall were observed over hill regions including Jammu& Kashmir, Himachal Pradesh, and Arunachal Pradesh (Fig. 7d). The temperature (Fig. 7g) has shown a significant increasing trend over northwest and north-east India whereas surface solar radiation (Fig. 7m) shows a decreasing trend in south peninsular India during the pre-monsoon season. The pre-monsoon relative humidity (Fig. 7j) has shown an increasing trend in some of the sub-divisions in south India (Kerala, Coastal Karnataka, South Interior Karnataka, Telangana) and north India (Uttar Pradesh, Bihar, and Sikkim). The analysis does not observe significant trends in pre-monsoon wind speed while decreasing trends observed over Kerala, Jharkhand, and NMMT sub-divisions (Fig. 7p).

In the case of monsoon season, similar (increasing) trends like Ea has observed in soil moisture (Fig. 7b) & RH (Rajasthan; Fig. 7k)), and temperature (Tamil Nadu, Kerala, Coastal Karnataka, SI Karnataka, east UP, Bihar, Jharkhand, and Gangetic West Bengal; Fig. 7h) during the study period. However, the net surface solar radiation has shown significant decreasing trends over northwest India and southeast India (Fig. 7n). The rainfall has shown decreasing trend in east Uttar Pradesh, Bihar, and Assam sub-divisions from 1980 to 2018 (Fig. 7e).

The trends in climatic parameters during the post-monsoon season shows that the RH has similar (increasing) trends similar to Ea over the sub-divisions located in foothills of the Himalayas (Uttar Pradesh, Bihar, and Sikkim) (Fig. 7l), whereas the rainfall shows inverse trends over Jammu &Kashmir, Uttar Pradesh, and Arunachal Pradesh (Fig. 7f). The trend analysis also observed decreasing trends in net surface radiation over the southwest coast, Tamilnadu, and east India (Fig. 7o). The surface temperature has shown an increasing trend over most of sub-divisions India except East India, north interior Karnataka, Vidarbha, and Jammu& Kashmir (Fig. 7i).

4.8 Relationship between Ea and Climatic parameters

Correlation analysis was carried out between the actual evaporation and the climatic parameters for pre-monsoon, monsoon, and post-monsoon seasons during the study period (Fig. 8). The correlation coefficients show that the actual evaporation has a strong (99% significant level) positive correlation with rainfall (Fig. 8d), soil moisture (Fig. 8a), and relative humidity (Fig. 8j), opposite correlation with SSR (Fig. 8p) over most of the sub-divisions in India (except northeast India, Jammu & Kashmir, and Himachal Pradesh) during pre-monsoon season. However, temperature (Fig. 8g) showed negative correlations over most of the sub-divisions and positive correlations over Jammu& Kashmir and Arunachal Pradesh during the pre-monsoon season.

The analysis observed a complex relationship with climatic parameters during the monsoon season. In general, rainfall (Fig. 8e), soil moisture (Fig. 8b), RH (Fig. 8k) have a positive correlation over the sub-divisions located in water-limited areas (i.e. South and north-west India) and a negative correlation found over energy-limited areas located in east India. In contrast to the rainfall, net surface radiation shows a significant positive correlation over energy-limited areas and a negative correlation over water-limited areas in India (Fig. 8q). Wind speed has shown a strong negative correlation over the arid and semi-arid regions of India (Fig. 8n).

In the case of the post-monsoon season also soil moisture, rainfall, and humidity showed a significant positive correlation with Ea over most of the sub-divisions in India except foothills of Himalaya regions (Fig. 8c, 8f, 8l). Net surface radiation shows a strong negative correlation over most of the subdivisions in India whereas a positive correlation is observed over western Himalaya (Jammu &Kashmir and Himachal Pradesh) and Arunachal Pradesh (Fig. 8r).

5 Discussion

The present study utilized the high-resolution gridded land evaporation data (Ep, Ea, Et, Eb, Ei, and Eo) from the Global Land Evaporation Amsterdam Model (GLEAM) to study the evapotranspiration trends over different meteorological sub-divisions in India during the period 1980–2018. The advantage of this data product is that the model algorithm involves long-term, consistent soil moisture and vegetation optical depth data which was observed from multiple microwave satellite sensors. The key findings of our study are that all India annual Ea has shown an increasing trend due to the tremendous increment in transpiration (vegetation growth) over most of the sub-divisions during the study period. Our results are consistent with the recent studies where they also observed a significant greening trend in India (Sarmah, Jia, and Zhang, 2018; Bhimala et al., 2020) due to the substantial improvement in the irrigated agricultural area, micro-irrigation methods, and precision agriculture (Jain, Kishore, and Singh, 2019).

The sub-division-wise trends reveal that annual Ea trends are prominent over Indo-Gangetic plains, arid and semi-arid regions located in north-west India and south India. However, rainfall has not shown any significant trends over these regions during the summer and winter monsoon seasons. This attribute that the greening trend (transpiration growth) in these areas is mostly dominated by the surface or groundwater irrigation compared to the rainfall. However, a few studies already confirmed that there is a significant increasing trend (29–63% between 1950-51 and 2014-15) in groundwater irrigation (well or tube well) and micro-irrigation over the water-limited regions like Rajasthan, Gujarat, Haryana, Andhra Pradesh, Tamilnadu and Karnataka (Jain, Kishore, and Singh, 2019). This intensive irrigation and subsequent croplands expansion played a pivotal role in the greening trend and transpiration growth over north-west and south India. Our study also demonstrated that there is a significant correlation (99% significance level) between soil moisture and Ea over north-west and south India during pre-monsoon, monsoon, and post-monsoon season during the study period.

In the case of the sub-divisions located in Indo-Gangetic regions also most of the croplands are dominated by surface and groundwater irrigation. Recently, Ambika and Mishra (2019) also found that the vegetation health improved over irrigated regions compared to the non-irrigated region of indo-Gangetic plains. A few studies also observed an increase in humidity and extreme moist heat stress due to the intensive irrigation over Indo-Gangetic regions in India (Mishra et al., 2020; Krakauer et al., 2020). Our study confirms that the increase in actual evaporation is due to the increase in transpiration component because the bare-soil evaporation shows a decreasing trend in Indo-Gangetic regions of India. Our study is also consistent with the earlier studies that the actual evaporation was positively correlated with mean air temperature & SSR (Krakauer et al., 2020) and negatively correlated with precipitation over Indo-Gangetic regions during the monsoon season. Our results attribute that the transpiration anomalies are controlled by the net surface solar radiation compared to the soil moisture because water is not a limiting factor in Indo-Gangetic regions.

6. Conclusions

The primary focus of this study is to characterize the annual and seasonal trends in actual evaporation and its components over different meteorological subdivisions using the satellite-based GLEAM gridded time-series data product. The advantage of the GLEAM data is that the evaporation stress factor is calculated based on the satellite-derived vegetation optical depth and the soil moisture data to improve the estimation of the actual evaporation at higher spatial resolution. The study found an increasing trend (1.3mm/year) in all India annual Ea during the period 1980–2018. Seasonal analysis reveals that the increasing trend is prominent during pre-monsoon (0.54mm/year) and monsoon season (0.52mm/year). To understand the drivers behind the increasing trend in Ea despite the fact the insignificant trends in potential evaporation and rainfall, the components of Ea (transpiration, bare soil evaporation, interception loss, and open water evaporation) were analyzed for annual and seasonal scales. Our analysis found that the annual transpiration has shown a significant increasing trend (1.9mm/year) due to the improvement in vegetation over India. However, the actual evaporation growth is less than transpiration due to the decreasing trend (-0.53mm/year) in bare soil evaporation in India.

Seasonal trends in Ea over different meteorological sub-divisions in India show that the more number of sub-divisions in arid and semi-arid regions (water-limited areas in south India and north-west India) shown an increasing trend in Ea during pre-monsoon and monsoon season. The analysis also found the increasing trend in Ea over the sub-divisions located in Indo-Gangetic plains (Uttar Pradesh, Bihar, Jharkhand, and Gangetic West Bengal) during the monsoon season. However, the transpiration and interception loss showed an increasing trend in most of the sub-divisions, actual evaporation increased only a few sub-divisions due to the decreasing trend in bare soil evaporation and open water evaporation. During pre-monsoon and post-monsoon seasons, the Ea has shown an increasing trend in hilly regions like Jammu & Kashmir due to the increasing trend in transpiration, bare soil evaporation, and open water evaporation.

Seasonal trends in soil moisture and rainfall show that the soil moisture has shown an increasing trend in south India and north-west India during the pre-monsoon season whereas the rainfall has shown an increasing trend in the sub-divisions located in the Western Ghats and South Karnataka and Andhra Pradesh. The analysis also observed an increasing trend in soil moisture over Rajasthan, Himachal, and Uttarakhand region during the monsoon season. Hence, the increasing trend in transpiration and interception loss over most of the sub-divisions in India may be due to the micro-irrigation (drip and sprinkler irrigation) methods followed by the modern agriculture system even though there are no significant trends in rainfall during monsoon or pre-monsoon season.

The relationship between actual evaporation and climatic factors shows that Soil moisture, Rainfall, and Relative humidity have a strong positive correlation whereas the temperature, wind speed, and surface shortwave radiation showed a strong negative correlation over most of the sub-divisions except northeast India. During monsoon season, the water-limited areas (south and north-west India) showed a positive correlation with soil moisture and relative humidity and negative correlation with wind speed and net surface shortwave radiation, whereas in energy-limited areas (east India), rainfall showed negative correlation and surface short wave radiation shown a positive correlation. Further, the modelling studies (Fowler, Pritchard, and Kooperman, 2018) are required to quantify the impact of irrigation and cropland expansion on evapotranspiration trends and precipitation recycling over India.

Statements & Declarations

Acknowledgments:

The authors are grateful to the Global Land Evaporation Amsterdam Model (GLEAM) for providing the terrestrial Evaporation data product (https://www.gleam.eu/). The authors acknowledge the IMD (India Meteorological Department), ESA-CCI for providing the rainfall and soil moisture data (https://cdsp.imdpune.gov.in/home_gridded_data.php; https://esa-soilmoisture-cci.org/ ). The authors also acknowledge the support and encouragement of the Head, CSIR 4PI, and Director, AcSIR.

Funding:

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Competing Interests:

The authors have no relevant financial or non-financial interests to disclose.

Author Contributions:

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Kantha Rao Bhimala and Raghavendra Pasad Kanike.

Kantha Rao Bhimala contributed to Conceptualization, Formal analysis, Investigation, Methodology, Writing Original draft, Validation, review and editing, manuscript administration, Visualization, Supervision.

Raghavendra Prasad K contributed to the data analysis, visualization, and investigation.

G K Patra contributed to the data analysis and visualization. 

Himesh S contributed to the editing of manuscript and investigation.

All authors read and approved the final manuscript.

Data Availability:  The datasets analyzed in the current study are available base on the request (https://www.gleam.eu/#downloads ). 

Ethics approval/declarations: Not Applicable

Consent to participate: Not Applicable

Consent for publication: Not Applicable

References

  1. Allen RG, Pereira LS, Raes D, Smith M (1998) FAO Irrigation and drainage paper No. 56. Rome: Food and Agriculture Organization of the United Nations 56(97):e156
  2. Ambika AK, Mishra V (2019) Observational evidence of irrigation influence on vegetation health and land surface temperature in India. Geophys Res Lett 46(22):13441–13451
  3. Ambika AK, Wardlow B, Mishra V (2016) Remotely sensed high resolution irrigated area mapping in India for 2000 to 2015. Sci Data 3(1):1–14
  4. Bandyopadhyay A, Bhadra A, Raghuwanshi NS, Singh R (2009) Temporal trends in estimates of reference evapotranspiration over India. J Hydrol Eng 14(5):508–515
  5. Bhimala et al (2020) Identification of vegetation responses to soil moisture, rainfall, and LULC over different meteorological subdivisions in India using remote sensing data. Theor Appl Climatol 142(3):987–1001
  6. Budakoti S, Chauhan T, Murtugudde R, Karmakar S, Ghosh S (2021) Feedback from Vegetation to Interannual Variations of Indian Summer Monsoon Rainfall. Water Resour Res 57(5):e2020WR028750
  7. Chattopadhyay N, Hulme M (1997) Evaporation and potential evapotranspiration in India under conditions of recent and future climate change. Agric For Meteorol 87(1):55–73
  8. Chaudhuri S, Dutta D (2014) Mann–Kendall trend of pollutants, temperature and humidity over an urban station of India with forecast verification using different ARIMA models. Environ Monit Assess 186(8):4719–4742
  9. Dorigo W, Wagner W, Albergel C, Albrecht F, Balsamo G, Brocca L, Lecomte P (2017) ESA CCI Soil Moisture for improved Earth system understanding: State-of-the art and future directions. Remote Sens Environ 203:185–215
  10. Fowler MD, Pritchard MS, Kooperman GJ (2018) Assessing the impact of Indian irrigation on precipitation in the irrigation-enabled community earth system model. J Hydrometeorol 19(2):427–443
  11. Gilbert RO (1987) Statistical methods for environmental pollution monitoring. John Wiley & Sons
  12. Gocic M, Trajkovic S (2013) Analysis of changes in meteorological variables using Mann-Kendall and Sen's slope estimator statistical tests in Serbia. Glob Planet Change 100:172–182
  13. Goroshi S, Pradhan R, Singh RP, Singh KK, Parihar JS (2017) Trend analysis of evapotranspiration over India: observed from long-term satellite measurements. J Earth Syst Sci 126(8):1–21
  14. Gruber A, Scanlon T, van der Schalie R, Wagner W, Dorigo W (2019) Evolution of the ESA CCI Soil Moisture climate data records and their underlying merging methodology. Earth Syst Sci Data 11(2):717–739
  15. Hobbins MT, Ramírez JA, Brown TC (2004) Trends in pan evaporation and actual evapotranspiration across the conterminous US: Paradoxical or complementary? Geophys.Res. Lett31(13)
  16. Jain R, Kishore P, Singh DK (2019) Irrigation in India: Status, challenges and options
  17. Jaswal AK, Rao GP, De US (2008) Spatial and temporal characteristics of evaporation trends over India during 1971–2000. Mausam 59(2):149–158
  18. Jung et al (2010) Recent decline in the global land evapotranspiration trend due to limited moisture supply. Nature 467(7318):951–954
  19. Kantharao B, Rakesh V (2018) Observational evidence for the relationship between spring soil moisture and June rainfall over the Indian region. Theor Appl Climatol 132(3):835–849
  20. Kaur S, Diwakar SK, Das AK (2017) Long term rainfall trend over meteorological sub divisions and districts of India. Mausam 68(3):439–450
  21. Kelkar RR, Sreejith OP (2020) Meteorological sub-divisions of India and their geopolitical evolution from 1875 to 2020. Mausam 71(4):571–584
  22. Konapala G, Mishra AK, Wada Y, Mann ME (2020) Climate change will affect global water availability through compounding changes in seasonal precipitation and evaporation. Nat Commun 11(1):1–10
  23. Krakauer NY, Cook BI, Puma MJ (2020) Effect of irrigation on humid heat extremes. Environ Res Lett 15(9):094010
  24. Martens B, Miralles DG, Lievens H, Van Der Schalie R, De Jeu RA, Fernández-Prieto D, Verhoest NE (2017) GLEAM v3: Satellite-based land evaporation and root-zone soil moisture. Geosci Model Dev 10(5):1903–1925
  25. Martens et al (2018) Terrestrial evaporation response to modes of climate variability. npj clim atmos sci 1(1):1–7
  26. Miralles DG, De Jeu RAM, Gash JH, Holmes TRH, Dolman AJ (2011) Magnitude and variability of land evaporation and its components at the global scale. Hydrol Earth Syst Sci 15(3):967–981
  27. Mishra V, Ambika AK, Asoka A, Aadhar S, Buzan J, Kumar R, Huber M (2020) Moist heat stress extremes in India enhanced by irrigation. Nat Geosci 13(11):722–728
  28. Niyogi D, Kishtawal C, Tripathi S, Govindaraju RS (2010) Observational evidence that agricultural intensification and land use change may be reducing the Indian summer monsoon rainfall.Water Resour. Res46(3)
  29. Oki T, Kanae S (2006) Global hydrological cycles and world water resources. Science 313(5790):1068–1072
  30. Padmakumari B, Jaswal AK, Goswami BN (2013) Decrease in evaporation over the Indian monsoon region: implication on regional hydrological. cycle Clim Change 121(4):787–799
  31. Pai et al (2014) Development of a new high spatial resolution (0.25° X 0.25°) Long period (1901–2010) daily gridded rainfall data set over India and its comparison with existing data sets over the region. Mausam 65(1):1–18
  32. Paul S, Ghosh S, Oglesby R, Pathak A, Chandrasekharan A, Ramsankaran RAAJ (2016) Weakening of Indian summer monsoon rainfall due to changes in land use land cover. Sci Rep 6(1):1–10
  33. Phuong DND, Tram VNQ, Nhat TT, Ly TD, Loi NK (2020) Hydro-meteorological trend analysis using the Mann-Kendall and innovative-Şen methodologies: a case study. Int J Glob Warm 20(2):145–164
  34. Rehana S, Monish NT (2020) Characterization of regional drought over water and energy limited zones of India using potential and actual evapotranspiration. Earth Space Sci 7(10):e2020EA001264
  35. Rehana S, Monish NT (2021) Impact of potential and actual evapotranspiration on drought phenomena over water and energy-limited regions. Theor Appl Climatol 144(1):215–238
  36. Roderick ML, Hobbins MT, Farquhar GD (2009) Pan evaporation trends and the terrestrial water balance. II. Energy balance and interpretation. Geogr Compass 3(2):761–780
  37. Sarmah S, Jia G, Zhang A (2018) Satellite view of seasonal greenness trends and controls in South Asia. Environ Res Lett 13(3):034026
  38. Soni A, Syed TH (2021) Analysis of variations and controls of evapotranspiration over major Indian River Basins (1982–2014). Sci Total Environ 754:141892
  39. Trenberth KE, Fasullo JT, Kiehl J (2009) Earth's global energy budget. Bull Am Meteorol Soc 90(3):311–324
  40. Wilks DS (2011) Statistical methods in the atmospheric sciences. 100 Academic press
  41. Yadav S, Deb P, Kumar S, Pandey V, Pandey PK (2016) Trends in major and minor meteorological variables and their influence on reference evapotranspiration for mid Himalayan region at east Sikkim, India. J Mt Sci 13(2):302–315
  42. Zeng Z, Piao S, Lin X, Yin G, Peng S, Ciais P, Myneni RB (2012) Global evapotranspiration over the past three decades: estimation based on the water balance equation combined with empirical models. Environ Res Lett 7(1):014026
  43. Zhan S, Song C, Wang J, Sheng Y, Quan J (2019) A global assessment of terrestrial evapotranspiration increase due to surface water area change. Earths Future 7(3):266–282
  44. Zhang Y, Chiew FH, Peña-Arancibia J, Sun F, Li H, Leuning R (2017) Global variation of transpiration and soil evaporation and the role of their major climate drivers. J Geophys Res Atmos 122(13):6868–6881
  45. Zhang Y, Peña-Arancibia JL, McVicar TR, Chiew FH, Vaze J, Liu C, Pan M (2016) Multi-decadal trends in global terrestrial evapotranspiration and its components. Sci Rep 6(1):1–12
  46. Zotarelli L, Dukes MD, Romero CC, Migliaccio KW, Morgan KT (2010) Step by step calculation of the Penman-Monteith Evapotranspiration (FAO-56 Method).Inst. Food Agric. Sci., Univ. Fla 28