Dual Nature of Land Ocean Thermal Contrast during Pre-Monsoon and Onset Phase of Indian Summer Monsoon

Abstract

Under faster Indian Ocean (IO) warming, several thermodynamical properties of the atmosphere over the Indian sub-continent change abruptly. The present study has evaluated the temperature field using ERA5 and IMDAA reanalysis. From the climatological evolution, it has been observed that before monsoon onset over Kerala (MOK) not only does meridional tropospheric temperature gradient reverses from negative to positive but the surface LOTC also decline very sharply. Interannual variation of LOTC shows that there is no significant trend, however, warming since 1980 may lead to an increase in variability. The reason behind having no trend in LOTC may be attributed to land ocean warming ratio (WR). Composite analysis depicts that except for early MOK, surface LOTC decreases sharply before MOK while deep tropospheric LOTC or meridional tropospheric temperature increases. The climatological average of pre-monsoonal average is about −5.52 K which has been found slightly higher (−4.9 K) during the early MOK years and found slightly lower (−5.7 K) during the delayed MOK years. Hence deep tropospheric LOTC is mostly used to identify the onset of MOK while surface LOTC can be utilized to predict MOK. However, to make a more precise MOK prediction, the interaction between three-dimensional temperature field with large-scale flow needs to explore.

1. Introduction

India being home to a large section of the agriculture-dependent population, thrives largely on the Indian Summer Monsoon Rainfall (ISMR). Besides deciding the agricultural output of a year, the onset of ISMR plays a pivotal role in the transition of the weather from the scorching summer season to a rainy/monsoon season. More than a billion people in South Asia are impacted by ISMR, which accounts for approximately 80% of the total annual rainfall from June to September. The onset of ISMR is marked by the monsoon onset over Kerala (MOK) in early June followed by a northward propagation of the monsoon winds and thereby covering the whole country by mid-July. The onset and advancement of the monsoon are primarily attributed to the differential heating of land and ocean that results in the formation of a gradient between the land surface temperature and ocean surface temperature [1, 2, 3, 4], often known as the Land Ocean Thermal Contrast (LOTC). LOTC is usually expressed by an amplification factor or a ratio factor, which is defined as the relationship between the change in land temperature to the change in ocean temperature. From various observations and simulations of the general circulation model (GCM), it has been seen that the global amplification factor has a value of 1.5 approximately [5, 6, 7]. However, the factor also undergoes variation depending on the latitude. In the multi-model average, the local minimum in the tropics is 1.2, and the maximum in the subtropics is 1.6 [8].

The role behind the land ocean thermal contrast being the primary driver of monsoon was well explained by Sir Edmund Halley in the seventeenth century. During the summer season, the land gets heated up faster than the oceans because of which a low-pressure region is created over the land mass followed by the formation of a high-pressure region over the oceans. Also, the polar shift of the Inter Tropical Convergence Zone (ITCZ) to the position over South Asia strengthens the center of the thermally induced low-pressure center. This creates a pressure gradient from the ocean to the land, causing warm and humid air to flow from the ocean to the land. The situation is exactly the opposite during the winter season. The landmass cools faster than the oceans during the winter season. So, a high-pressure area is formed over the land and a low-pressure region is established over the ocean. This results in the establishment of the land-to-ocean pressure gradient followed by the outflow of cold and dry air from the land to the ocean. The advancement of the monsoon winds northward implies a shift in the deep convection from the equatorial to continental regions [3, 9, 10].

As stated earlier, the monsoon onset is mostly thought to be due to the LOTC. The classic theory of LOTC only takes account of the surface temperature and can only lead to shallow circulation. However, the meridional gradient in the upper troposphere is directly proportional to the meridional gradient of the deep troposphere heating, which can accelerate the deep tropospheric circulation [11]. The upper levels show an appreciable difference between the temperatures of the pre-monsoon and the monsoon months [12]. He et al. [13] through their work suggested that a considerable amount of tropospheric warming starts in May over the areas stretching from the Eastern Tibetan Plateau (TP) to the South China Plain. As a result, the meridional temperature gradient on south of the eastern Plateau reverses and the low-level south-westerlies start blowing over the Bay of Bengal. Then, the onset of the Indian monsoon occurs in June with the increase of temperature over a wide area covering Saudi Arabia, Iran, Afghanistan and the western Plateau. The thermal effect of TP is quite prominent as the air over it, is always warmer than its surrounding areas [14]. Flohn [15, 16] proposed that seasonal warming of the upper surface of the TP and subsequent changes in meridional temperature gradient and pressure gradient to the south of latitude 35°N caused large-scale changes in Asian atmospheric circulation and outbreaks of monsoon over the Indian subcontinent. Li and Yanai [17] confirmed and concluded that the change in the meridional temperature gradient is due to the increase in temperature over the Tibetan Plateau, and there is no obvious temperature change on the Indian Ocean (IO), indicating that in the spring month of the year, TP acts as an independent heat source, separate from the more intense heat sources that are associated with the equatorial IO rain belt. They further noted that when, on one hand, the sensible heating of the surface leads to the Tibetan heat source, the release of latent heat of condensation contributes to the oceanic heat source and it is the sensible heating of the land surface that is responsible for the reversal of the meridional temperature gradient than the latent heat of condensation. This is because though the intensity of latent heat of condensation may be high, due to the adiabatic cooling, it is unable to produce any significant temperature change over the IO. Hence LOTC may get controlled by temperature change over land during onset. After the onset, ISMR has a northward propagation and moves from the southern part of the country to cover the whole country. The progress of ISMR over India is perpendicular to the mean winds blowing from the southwest from June to September [18]. After the onset of ISMR, there is a significant decrease in land temperature thereby resulting in a cooling of the land surface. However, the latent heat released from the convection phenomenon keeps the troposphere above the land warm thereby maintaining the LOTC and consequently helping in the advancement of the monsoon throughout the country [19]. Hence, deep tropospheric LOTC is capable of driving winds that carry a massive amount of moisture from the warm oceans and can impact large-scale changes. The above discussion clears that surface LOTC and deep tropospheric LOTC have different aspects and effects during onset and advancement. Under the global warming scenario, it has been observed that the Land surface temperature (LST) increases more rapidly than the ocean surface temperature (OST) in the northern hemisphere under increasing greenhouse gasses. Although through literature review may provide insight about the dependency of LOTC on land temperature, it is not clear how does LOTC modify in response to change in land and ocean temperature. To understand, how does thermal contrast impact ISM, studies are mostly focused on surface temperature rather than deep tropospheric temperature. Jin and Wang [20] showed that ISM has been started reviving since 2002 under favorable surface LOTC while the theory of association of surface LOTC with surface LOTC has been proved to be null and void by Gadgil et al. [21]. However, how the deep tropospheric LOTC evolves and effect Indian monsoon onset and advancement is not well explored.

In this present study, LOTC at the surface and for the deep troposphere has been evaluated and explored in detail. A brief climatology of surface temperature and deep tropospheric temperature along with monthly change and variation at an inter-annual scale has been assessed to understand the spatial pattern of LOTC evolution during ISM onset and advancement. The present study has also explored the sensitivity of LOTC towards the surface and deep tropospheric temperature for both land and ocean. Afterward, the study has tried to explore the effect of LOTC during different types of ISM onsets. Section 2 describes about data and methodology used in the present study. Section 3 includes the detailed observed analysis followed by Section 4 which summarizes and provides a brief conclusion.

2. Data and methodology

2.1 Data

To study the impact of LOTC on the onset and advancement of ISMR the analysis has been carried out for five months - March, April, May, June and July using long-term data. Parameters considered for the purpose of study are - Temperature at 2 m, Temperature at pressure levels, Precipitation. All meteorological data except precipitation used for evaluation have been taken from the fifth-generation reanalysis datasets of the European Centre for Medium-Range Weather forecasts (ERA5;[22]) and the Indian Monsoon Data Assimilation and Analysis (IMDAA; [23]) reanalysis. From both sources, data has been collected at monthly and hourly time scales for all available time period. The evaluation has been carried out using two datasets (IMDAA and ERA5) so as to check the robustness of the data and draw a stronger conclusion from the results obtained. ERA5 provides all data at 25 km spatial resolution from 1959 to 2021 and IMDAA provides data at a horizontal resolution of 0.12° × 0.12° from 1979 to 2021. For each case, pressure level temperature has been taken for 600 hPa, 500 hPa, 400 hPa, 300 hPa and 200 hPa. Hourly data from both sources have been converted into daily by averaging. Tropospheric Temperature has been calculated by averaging temperature from 600 hPa to 200 hPa as per [11].

2.2 Methodology

The climatological and monthly tendency has been analyzed to understand the change in the temperature field and development of LOTC. Through this analysis, the study has obtained a domain over both land and ocean to study LOTC. A spatial linear trend analysis has been performed. Since the linear trend does not have any statistical significance, a Man-Kendall trend analysis has been implemented along with linear trend analysis for area-averaged data. To explore the variability of LOTC, two parameters have been evaluated. First is Land Ocean thermal contrast which is the standardized difference between the land temperature and the ocean temperature. The positive value of LOTC signifies higher land temperature. Another parameter is land ocean warming ratio (WR) which is the ratio of change in land temperature and change in ocean temperature. WR provides insight about into the direction in which land and ocean change simultaneously. The positive value of WR signifies that land and ocean temperature changes are in phase i.e. either two are increasing or decreasing. The negative value of WR depicts out-phase change in both temperature fields. For further analysis, monsoon onset over Kerala (MOK) has been classified as early, delay and normal MOK considering mean MOK as 1st June with a standard deviation of 8 days. Thereafter a composite analysis has been performed to analyze LOTC field during different MOK.

3. Results

3.1 Climatological analysis

Pre-monsoon and pre-onset climatology of the temperature field has been analyzed using all available data from both data sources. The climatological evolution of the Surface temperature at 2 m (Figure 1) from both sources - IMDAA and ERA5 show similar results. The graphs show a relatively low temperature over northern India and the Tibetan Plateau (TP) region compared to the rest of India for all the months from March to July. However, it can be observed that as the season changes from the pre-monsoon period to the monsoon period the temperature over the TP increases but remains low than the rest of the Indian Subcontinent. Also, the temperature over the Indian landmass is seen to increase by very small amounts from March to July. The surface temperature over the Indian Ocean remains warm from March to July only varying by small amounts. Tropospheric temperature (TT) over the northern part of India and the Tibetan Plateau shows an increase from March to July in Figure 2. The increase is small during March, April and May which intensify in the month of June and July. In July, the tropospheric temperature is high in the Tibetan plateau region thereby helping the monsoon advancement. The IMDAA data shows a less amount of warming over the rest of India compared to the ERA 5 data. The TT over the Indian Ocean is also seen to increase from March to July and similar to the land temperature IMDAA shows a lesser temperature than that shown by the ERA 5 data.

Figure 1.

Climatology of surface temperature at 2 m during March, April, May, June and July from (a) IMDAA reanalysis and (b) ERA5 reanalysis.

Figure 2.

Climatology of tropospheric temperature during March, April, May, June and July from (a) IMDAA reanalysis and (b) ERA5 reanalysis.

The monthly tendency plots (Figure 3) show that the intensity of warming up of the land surface is increasing during the pre-monsoon period. The land surface of the Indian subcontinent and Tibetan plateau region is warmer in April and May compared to March and April, respectively. However, a decrease in land surface temperature is seen in Southern India in June compared to May. This decrease in land temperature shifts from Southern India to the Central part of India and so a decline of surface temperature over land is seen in this part from July to June. A similar pattern of tendency is seen in the Tropical Indian Ocean. The warming of the ocean surface temperature increases in April and May compared to March and April, respectively. On the other hand, there is a decrease in the temperature during June and July compared to May and June, respectively. From the graphs, it can be concluded that the warming intensity and the change in temperature are more on land than that in the ocean. The monthly tendency plots of the TT (Figure 4) show that over the land the tropospheric temperature warming intensity increases by a small amount from March to April, and increases significantly over the Northern India and Tibetan Plateau region during the onset and advancement of the ISMR. But, the amount of increase of the warming decreases during the advancement than during the onset phase. On the other hand, the tropospheric temperature over the Indian ocean shows a positive tendency in April and May compared to March and April, respectively. However, the temperature over the TIO starts decreasing in June and July when compared to the Temperature in May and June, respectively. Both sources (IMDAA and ERA5) show nearly the same results but the tendency of tropospheric temperature over the Indian Ocean is slightly different for the July–June plot where the ERA5 data shows a larger decrease in the tropospheric temperature of the Indian Ocean than that in IMDAA data.

Figure 3.

Climatological monthly tendency of surface temperature at 2 m during pre-monsoon and onset phase of ISM evaluated from (a) IMDAA reanalysis and (b) ERA5 reanalysis.

Figure 4.

Climatological monthly tendency of tropospheric temperature during pre-monsoon and onset phase of ISM evaluated from (a) IMDAA reanalysis and (b) ERA5 reanalysis.

Based on the above analysis, it has been observed that thermal contrast regions are not the same for the surface and deep troposphere. Surface temperature mostly changes due to the radiative effect and due to the contrast of heat capacity between land and ocean whereas the temperature of the deep troposphere is mostly controlled by latent heating. The above analysis shows that before the onset of the monsoon, a small region of the southern peninsular develops LOTC near the Kerala region where deep tropospheric LOTC becomes prominent and a significant structure during the onset of monsoon. Hence for further land analysis of LOTC, land and ocean areas have been identified separately for surface and deep troposphere. For surface land and ocean, area is bounded by 8-18 N;78-80E and 10-20 N;60-75E, respectively, and for the deep troposphere, land and ocean area are bounded by 25-35 N;60-100E and 10-5 N; 50-95E, respectively (Figure 5). Figure 6 shows the climatological evolution of LOTC from March to July using the IMDAA reanalysis data and ERA5 reanalysis data. The y-axis on the left and right sides shows the surface LOTC plotted in black and deep tropospheric LOTC plotted in green respectively. From the graphs of both sources, it is evident that during the pre-monsoon period, surface LOTC decreases and increases with the onset of the monsoon. While the deep tropospheric LOTC shows almost a linear increase during the pre-monsoon and the monsoon periods. The 0 line is plotted for the deep tropospheric LOTC as it shows a negative value during the pre-monsoon period, however, the surface LOTC has a positive value through the 5 monthsperiod. It has been observed that not only meridional tropospheric temperature gradient becomes positive but also the surface land ocean thermal contrast shows a sharp decline just before onset (Figures 7 and 8).

Figure 5.

Land Ocean area for surface temperature (green box) and tropospheric temperature (red box).

Figure 6.

Climatological evolution of Land Ocean thermal contrast starting from March to July evaluated from (a) IMDAA reanalysis and (b) ERA5 reanalysis. Left side y-axis signify surface LOTC (black line) and right-side y-axis signify deep tropospheric LOTC (green line). “0” line is marked against deep tropospheric LOTC.

Figure 7.

Interannual variation of LOTC at the surface (left panel) and in the deep troposphere (right panel). Man-Kendall trend analysis has been marked with its statistical significance. Trend value significant at 95% level has been marked in bold.

Figure 8.

Monthly climatology of Land Ocean warming ratio.

3.2 Variability analysis

Interannual variation of LOTC at the surface level and in the deep troposphere has been attributed through trend analysis. It has been observed that there is no significant visual linear trend in LOTC at both levels. Man-Kendall trend analysis depicts that the surface LOTC shows a significant decreasing trend (IMDAA reanalysis) during May and a significant increasing trend (ERA5 reanalysis) during June and July. The deep troposphere LOTC interannual variation, on the other hand, shows a significant declining trend (ERA5 reanalysis) during June and July.

Change in LOTC may be attributed through changes in land and ocean temperature which can be understood through the warming ratio which provides insight into the direction of change. In the month of March positive value of WR for all data sources signifies that the direction of change is the same over both land and ocean. In the month of April, there is some inconsistency in the outcome. ERA5 data signifies that land and ocean temperature changes in the same direction while IMDAA signifies the opposite. In the month of May, except for the deep tropospheric, both data source signifies surface temperature changes over land and ocean in the same direction. For the month of June, all data depicts the change in temperature in the same way. However, if seasonal averages are considered, it may be stated that land and ocean both manifest changes in temperature in the same direction. Frequency analysis (Figure 9) suggests that more than 60% cases of land and ocean temperature changes are in phase.

Figure 9.

Frequency analysis of Land Ocean warming ratio.

3.3 Impact of Land-Ocean thermal contrast during onset and advancement of Indian summer monsoon

To understand the impact of LOTC on monsoon onset and advancement, the present study has mainly focused on the movement of convective activity. Northward propagation of convection attributed through an anomaly in OLR, is a significant trigger during ISM onset. The present study has analyzed and compared the LOTC field along with northward propagation for early, delay and normal MOK years. It has been observed that during early MOK, surface LOTC decreases gradually while during delay and normal onset, the declination LOTC is sharp. Also, Deep tropospheric LOTC evolves according to convective activity. Hence it is clear that to trigger MOK, surface LOTC is significant rather than deep LOTC. Deep tropospheric LOTC evolution is the manifestation of convective activity (Figure 10 and Table 1). It has been observed that during early MOK deep tropospheric LOTC is higher than normal during pre-monsoon season while during delay, it has been noted to be lower than normal. For surface LOTC, the variation in surface LOTC during early, delay and normal is not same (Figure 11).

Figure 10.

Composite analysis of northward propagation of convective activity averaged over 70E-80E and evolution of surface and deep tropospheric LOTC during early (left panel), delay (middle panel) and normal (right panel) MOK years. Left y-axis defines latitude (in degree N) and right y-axis defines LOTC (in degree K). Solid black color for ERA5 and dotted line for IMDAA reanalysis. The upper panel is for surface LOTC and the bottom panel for deep tropospheric LOTC. In both cases shaded color define OLR anomaly.

Figure 11.

Variation of land ocean thermal contrast during normal MOK, early MOK and delay MOK for (a) deep tropospheric LOTC and (b) surface LOTC.

Early	Delay	Normal
1960,1961,1962,1969,; 1985,1990,1999,2009,2013	1967,1972,1979,1983,1986,1; 995,1996,1997,2002,2003	1959,1963,1964,1965,1966,1968,1970,; 1971,1973,1974,1975,1976,1977,1978,; 1980,1981,1982,1984,1987,1988,1989,; 1991,1992,1993,1994,1998,2000,2001,; 2004,2005,2006,2007,2008,2010,2011,; 2012,2014,2015,2016,2017,2018,2019,; 2020,2021

Table 1.

Classification of early, delay and Normal monsoon onset over Kerala based on MOK mean date 1st June.

4. Conclusions

Land ocean thermal contrast has been the most debated topic in the field of Indian summer Monsoon studies. One group of scientists denied the theory of LOTC behind the Indian monsoon and another group provides justification of keeping the LOTC theory. The present study aims to understand the contradiction and explore a completely new insight about LOTC. Most of the study pertaining to ISM circulation and rainfall has been associated with surface LOTC while deep tropospheric LOTC has been associated with MOK.

The present study evaluated the temperature field at the surface and for the deep troposphere using the two most current reanalyzes i.e., ERA5 and IMDAA reanalysis. Although these two data have been at two different spatial resolutions, these two data can explicitly capture the climatological behavior. While comparing these two data for the temperature field, it has been observed that it has range differences with a similar spatial pattern. With both data sets, trend analysis has been performed which shows the disparity in results unless the trend is very prominent in both data in the same way. It has been assumed that IMDAA reanalysis resolved the Indian feature much more than ERA5 because of high resolution as IMDAA reanalysis has been only generated by focusing on the Indian sub-continent.

A detailed climatological analysis has been performed with the temperature field at monthly, daily and interannual scales. Spatial climatology and monthly tendency analysis explored that surface LOTC and deep tropospheric LOTC has different domain over land and ocean which has been identified and further analysis of LOTC over those areas has been evaluated. From the climatological evolution of surface and deep tropospheric LOTC, it has been observed that before MOK not only meridional tropospheric temperature gradient reverses from negative to positive but surface LOTC also decline very sharply. Interannual variation of LOTC shows that there is no significant trend, however, warming since 1980 may lead to an increase in variability. The reason behind having no trend in LOTC may be attributed through WR ratio which depicts that in more than 60% cases land and ocean changes are in phase.

The evolution of LOTC analysis depicts that except for early MOK, surface LOTC decreases sharply before MOK while deep tropospheric LOTC or meridional tropospheric temperature increase may be a manifestation of increasing convective activity. Hence deep tropospheric LOTC is mostly used to identify the onset of MOK while surface LOTC can be utilized to predict MOK.

Acknowledgments

The authors are also grateful to IntechOpen for giving the opportunity to contribute to this book and also thankful to the editor for accepting the proposal.

Conflict of interest

“The authors declare no conflict of interest.”