Spatiotemporal trends in sunshine hours over India during three decades from 1988 to 2018

Abstract

This study delves the trends of sunshine reaching the earth surface, both temporally and spatially, across nine geographically diverse regions including 20 stations in India, spanning the years 1988 to 2018. The monthly sunshine hours (SSH) analysis concluded significant increment from October to May followed by significant drops from June to July in six regions, except northern inland and Himalayan region that showed comparatively opposite monthly trends. The trend analysis depicted annual negative trend in all geographical regions with different rate (east coast: − 4.88 h/year; west coast: − 8.62 h/year; northern inland − 13.15 h/year; central inland: − 4.71 h/year; Deccan plateau: − 3.05 h/year; north eastern region: − 1.33 h/year; Himalayan region: − 9.47 h/year; island location Arabian Sea: − 5.72 h/year and Bay of Bengal − 6.10 h/year). Seasonal trends were found significant decline, but due to regional meteorological variation, accompanying Twomey effect may lead levelling off in SSH over north east region during monsoon and post-monsoon season. Analogous to annual SSH trend, intra-annual anomaly results were also depicting consistent decline in all geographical locations of India. The study reveals persistent decline of SSH in Indian subcontinent on all temporal scales excluding north eastern region where mild seasonal levelling off was found.

Introduction

Solar radiation is one of the most significant unconventional energy sources. It has emerged as one of the most promising strategies to mitigate ill effects of the conventional energy processes. Environmental degradation, energy security, climate change, compromised health, etc., can be addressed by such non-conventional energy generation methods. In recent decades, however, various regions across the globe have shown significant fluctuations in observed solar radiation. Stanhill & Cohen revealed raising concerns about the solar dimming and reported an average reduction of 0.51 W/m² or 2.7% per decade over the past 50 years¹. In the meantime, many studies have discussed the issue of dimming and brightening from various parts of the globe^2–8 (Table 1). A few studies have also reported decrease in solar radiation from various regions of India^37–39. Kumari et al. reported decreasing trends in solar radiation across India, with an average decrease of solar radiation by approximately 0.86 W/m² per year between 1981 and 2004²¹. Several studies in India reported a persistent solar dimming in 21st century^25,28,40. One of the most accepted reason for such dimming is transmissivity change of the earth’s atmosphere due to an increase in aerosol number density⁸. During the 1990s, India’s economic growth drove urbanization, land-use changes, and industrialization, leading to increased fossil fuel consumption, vehicular emissions, and biomass burning. These activities significantly raised aerosol concentrations in to the atmosphere and thus reduced solar radiation³⁹. Meanwhile, China and Japan adopted stricter air quality measures after the 1990s. China’s Clean Air Action Plan and Japan’s implementation of clean technologies and emission controls successfully lowered aerosol levels, leading to a brightening trend. While several studies have shown long term solar radiation fluctuations^15,26, sunshine hours (SSH) still remained relatively under-researched, particularly in India, although it can be understood that a negative change in its value may severely impact the solar energy generation stratagem. The solar radiation is defined as the amount of energy received by a surface area per unit of area and unit of time whereas, SSH is quantified as the duration for which the sun illuminates above 120 Wm⁻². Understanding the variations in SSH is crucial for atmospheric modelling and research on climate change dynamics⁴¹ as well as in understanding the Agricultural Meteorology and available solar energy.

Table 1.

Asian country solar dimming and brightening period.

Sunshine hours are dependent to the concurrent weather conditions that can obscure the solar incidence. Presence of cloud is one of the most prevalent ways that can obstruct the incoming sunlight, although such conditions can also be achieved by the presence of fog or dust storms. Hygroscopic aerosols, acting as cloud condensation nuclei (CCN), play important roles in modulation of cloud properties and hydrological cycle, including the life time of clouds. As per theory of condensation for warm cloud system, bigger cloud droplets are formed when natural balance of aerosols and humidity persists, and formation of perceptible clouds is achieved⁴². For higher number of aerosols in the same amount of humidity, however, a greater number of smaller size cloud droplets are developed and the cloud lifetime is prolonged. This is an important phenomenon in the atmosphere, known as the aerosol indirect effect^43,44. The synergic association between aerosols and atmospheric humidity actually defines the size of droplets. It is reported that, to a large extent, the condensation nuclei (CCN) and ice nuclei (IN) determine the cloud microstructure and, consequently, influence the dynamic response of clouds to aerosol-induced changes to the precipitation. In other words, it can be said that the aerosols are the seed for modulating the cloud residence time in the atmosphere. Consequently, it can be hypothesised that clouds, being the major atmospheric factor that can obscure the solar radiation reaching to the ground, would be impacted, and the hours during which the solar radiation can reach to the surface would also be impacted. Jnanesh et al. emphasized that aerosols influence cloud microphysics by modulating the cloud droplet size⁴². In another study, Lal et al. documented larger cloud droplets over a high-humidity area and a smaller size over a low-humidity area⁴⁵.

The present study analyses data from 1988 to 2018, encompassing both historical and recent datasets to examine SSH trends across India. Covering an extensive temporal scale and the entire spatial extent of the country, it offers a comprehensive evaluation of spatial and temporal variability. Jaswal nicely reported a wide point scale spatio-temporal variability for the 1970–2006 periods²². A few other studies addressing SSH variability in India have either covered shorter timeframes or limited spatial extents. This makes our study unique, as it provides a more comprehensive evaluation of SSH trends across a broader spatial and temporal spectrum. The study is designed for various climatological regions distributed in the tropical and subtropical latitudes over the Indian subcontinent. The analytical part of the study first presents the spatial variability in terms of intra-annual, monthly, seasonal and intra-annual anomaly, and then shows the strength and direction of trends for intra-annual, monthly, seasonal SSH for all the regions.

Research methodology

Study area

The National Climatic Data Centre (NCDC; http://ncdc.noaa.gov/) has historical weather records. Indian SSH data for the span of 1988 to 2018 has been extracted from this database. The downloaded SSH dataset has undergone rigorous consistency checks, ensuring a minimum of 10 years of uninterrupted data availability. As a result, a comprehensive dataset comprising 20 weather stations has been compiled for further analytical exploration. Figure 1 illustrates the geographical distribution of these selected stations, categorized into distinct regions for representative analysis. This study focuses on the monthly data from these delineated regions, enabling a macroscopic understanding of the monthly, intra-annual, seasonal and intra-annual anomaly patterns prevalent in India during the specified time frame. The seasonal behaviour of India’s climate is categorized into four different seasons, i.e. winter (January-February-December, JFD), pre-monsoon (March-April-May, MAM), monsoon (June-July-August-September, JJAS), and post-monsoon (October-November, ON).

Fig. 1.

Illustration of geographic distribution of 20 weather stations across India. Colours of symbols represent stations in a particular region. (this figure is made using shape file obtained from https://onlinemaps.surveyofindia.gov.in⁵⁴ with the help of open access tool QGIS 3.38 available at from https://qgis.org/download/⁵⁵).

Numerical analysis

In this study, a nonparametric method, Mann–Kendall test (MK test) and Modified Mann- Kendall test (MMK test) were used because it works better with non-normally distributed data, unaffected by the outliers, and possess ability to deal with the missing data values^46–48. The comparative analysis of MK test and MMK test results does not show any change in the SSH overall trends and rates at annual as well as seasonal scale, however the only change observed was in the p-value. Consequently, the results were interpreted based on the MK test statistics. Many researchers have applied and found that MK test has been useful in similar types of application^49–51. The trend magnitude was estimated using Sen’s slope estimator test⁵². A positive value of Sen’s slope indicates an upward trend and the negative value indicates a downward trend in the time series.

Results and discussion

Eastern Coast of India

The east coast of India, including the cities Chennai, Machilipatnam and Bhubaneswar, experiences a daily pattern of evening cloudiness and clearer mornings due to the land-sea breeze effect. The annual mean SSH was found to be 2244± 84 h/year and annual MK trend analysis reveals that SSH declined at the rate of -4.88 h/year for a span of 1988 to 2018, with intermittent undulations occurring over periods of approximately 2 to 3 years (Fig. 2a). The average monthly SSH in this region was around 187 ± 46 h/month, with maximum and minimum values 274 h/month and 72 h/month, respectively. Figure 3a demonstrates monthly mean SSH variability, showcasing a low deviation from the climatological mean. Monthly variation of SSH for the east-coastal region illustrated increasing values from January (230 ± 26 h) to April (253 ± 26 h) and then again from October (173 ± 4 h) to December (200 ± 21 h). Conversely, a gradual decrease was observed from May (238 ± 21 h) to August (139 ± 38 h). The stations situated along the eastern coast of the Indian subcontinent, positioned at different latitudes, also exhibit distinct local behaviours. Chennai remains relatively cloudy during October and November due to northeast monsoon, while the onset of summer monsoon clouds from June onwards impacts cloudiness over Chennai and Machilipatnam, but not over Bhubaneswar. The overall east coast region was portraying the decreasing trend of SSH from June (159 ± 39 h) onward minimum in July (128 ± 50 h) and further increasing from October (173 ± 4 h) to May (238 ± 21 h).

Fig. 2.

Annual variation in of SSH over different regions of India. The black line depicts the Sen’s Slope of SSH over the time.

Fig. 3.

Monthly variation of SSH over different regions of India, box representing 25–75% of data, horizontal line is median, diamond shape is the mean SSH and bold line is median value.

Across the east coast, maximum SSH is observed during the pre-monsoon seasons (740 ± 25 h respectively) and minimum in post-monsoon season (360 ± 24 h; Fig. 4a). MK trend and Sen’s slop analysis, as detailed in Table 3, reveals negative values at a significance level of ≥ 95% for winter, pre-monsoon and monsoon seasons, though, for post-monsoon MK trend (Zc: -0.61) and Sen’s slop (-0.29) was negative and insignificant (p > 0.05). Temporal variations of SSH as anomaly along the east coast from 2001 to 2018 are depicted in Fig. 5a. The brightest year was 2004, with an anomaly of 162 h, while the darkest year was 2010, with an anomaly of -203 h. Figure 5a illustrates that the SSH exhibited an overall declining trend with Sen’s slope of -4.88 h/year, demonstrating a region specific increase in cloud cover, particularly during the winter seasons. Jaiswal et al. reported a systematic decline in SSH at the eastern region stations⁵³. Soni et al. also reported a decline SSH trend from 1970 to 1990 in Chennai station²⁵.

Fig. 4.

The seasonal variation of SSH over different region of India. The dotted line defines the overall trend of SSH and shaded area is its standard deviation.

Table 3.

MK test analysis details for nine geographically different regions of India.

Region	Seasonal SSH	Trend significance	p-value	Zc	Sen’s Slope	Trend	α
East coast	Winter	Yes	0.00	− 3.73	− 2.72	Down	95%
	Pre-monsoon	No	0.15	− 1.42	− 0.81	Down	95%
	Monsoon	Yes	0.00	− 2.88	− 1.61	Down	99%
	Post-monsoon	No	0.54	− 0.61	− 0.29	Down	95%
	Annual	Yes	0.00	− 3.64	− 4.88	Down	95%
West coast	Winter	Yes	0.00	− 3.81	− 2.33	Down	99%
	Pre-monsoon	Yes	0.00	− 3.16	− 2.63	Down	95%
	Monsoon	No	0.29	− 1.05	− 1.05	Down	95%
	Post-monsoon	Yes	0.00	− 3.28	− 1.75	Down	95%
	Annual	Yes	0.00	− 3.67	− 8.62	Down	99%
Northern inland	Winter	Yes	0.00	− 4.89	− 4.83	Down	99%
	Pre-monsoon	Yes	0.00	− 4.25	− 2.69	Down	95%
	Monsoon	Yes	0.00	− 3.33	− 2.46	Down	95%
	Post-monsoon	Yes	0.00	− 3.23	− 2.15	Down	99%
	Annual	Yes	0.00	− 5.40	− 13.15	Down	99%
Inland central	Winter	Yes	0.00	− 3.43	− 2.02	Down	99%
	Pre-monsoon	Yes	0.00	− 4.19	− 2.49	Down	99%
	Monsoon	No	0.23	− 1.19	− 1.21	Down	99%
	Post-monsoon	No	0.39	− 0.85	− 0.50	Down	95%
	Annual	Yes	0.00	− 3.19	− 4.71	Down	99%
Deccan plateau	Winter	Yes	0.00	− 3.60	− 1.44	Down	95%
	Pre-monsoon	No	0.26	− 1.29	− 0.33	Down	95%
	Monsoon	No	0.19	− 1.29	− 1.71	Down	95%
	Post-monsoon	No	0.15	− 1.41	− 0.62	Down	95%
	Annual	Yes	0.00	− 3.19	− 3.05	Down	95%
North eastern	Winter	Yes	0.02	− 2.34	− 1.23	Down	95%
	Pre-monsoon	No	0.16	− 1.41	− 0.53	Down	95%
	Monsoon	No	0.48	0.70	0.13	Up	95%
	Post-monsoon	No	0.77	0.29	0	Up	95%
	Annual	No	0.07	− 0.07	− 1.33	Down	95%
Himalayan region	Winter	No	0.76	− 0.31	− 0.27	Down	95%
	Pre-monsoon	Yes	0.04	− 2.01	− 2.04	Down	95%
	Monsoon	Yes	0.02	− 2.36	− 3.12	Down	99%
	Post-monsoon	Yes	0.00	− 2.80	− 2.78	Down	99%
	Annual	Yes	0.00	− 3.16	− 9.47	Down	99%
Arabian Sea	Winter	Yes	0.03	− 2.14	− 1.35	Down	95%
	Pre-monsoon	Yes	0.06	− 1.84	− 1.51	Down	99%
	Monsoon	No	0.58	− 0.54	− 0.82	Down	95%
	Post-monsoon	No	0.22	− 1.22	− 0.89	Down	99%
	Annual	Yes	0.00	− 3.48	− 5.72	Down	99%
Bay of Bengal	Winter	Yes	0.00	− 3.08	− 3.11	Down	99%
	Pre-monsoon	Yes	0.01	− 2.56	− 2.25	Down	95%
	Monsoon	No	0.97	− 0.03	− 0.04	Down	95%
	Post-monsoon	No	0.13	− 1.51	− 0.57	Down	95%
	Annual	Yes	0.00	− 3.67	− 6.10	Down	95%

Fig. 5.

Inter-annual anomaly trend of SSH in different regions of India.

West Coast of India

The mean monthly SSH for the west-coastal region was 192 ± 60 h, with a maximum recorded value of approximately 290 h/month and a minimum of about 57 h/month. This region encompasses three stations: Thiruvananthapuram (173 ± 44 h/month), Goa (213 ± 76 h/month), and Mumbai (189 ± 71 h/month). The annual mean SSH found 2301 ± 108 h/year (Fig. 2b) and annual MK trend analysis reveals that SSH was declined at the rate of -8.62 h/year. However, an apparent decreasing trend was observed in the 3-year running mean analysis, indicating a gradual decline in SSH throughout the study period. Notably, from 1988 to 1995, there was a levelling off from 1997 to 2000 in SSH, followed by a continuous decline until 2007. December to May typically experiences minimal disturbances therefore higher magnitude of SSH (Fig. 3b), but with the onset of pre-monsoon cloudiness, SSH start decreases from April onwards, reaching its lowest point during the active monsoon months, particularly in July (83 ± 27 h, Fig. 4b). The monthly SSH along the west coast exhibit a distinct array, with minimum values recorded during the months of June to September 117 to 155 h. In contrast, maximum SSH values are observed from January to April, ranging between 253 and 230 h.

The maximum SSH (727 ± 30 h) was reported during the winter season and followed by pre-monsoon season (712 ± 40 h). Conversely, the minimum SSH value (398 ± 28 h) was observed during the post-monsoon seasons. West coast pre-monsoon season is characterized by high temperatures and aridity, leading to convective activities such as sandstorms, dust storms, and thunderstorms across various parts of the country⁵⁶. Across all stations along the west coast there was uniformity in exhibiting maximum SSH during the winter (665 ± 32; 834 ± 73; 681 ± 34 h, respectively), followed by pre-monsoon (585 ± 52; 788 ± 26; 762 ± 43 h, respectively), while minimum SSH occurred during the post-monsoon (292 ± 40; 471 ± 70; 423 ± 23 h, respectively, Table 2). MK trend analysis, as detailed in Table 3, reveals negative values at a significance level of ≥ 95% for winter (− 2.33), pre-monsoon (− 2.63) and post-monsoon (− 1.75) whereas monsoon season showed an insignificant negative trend (− 1.05). Soni et al. identified an insignificant negative trend in winter SSH. Additionally, an all-sky annual decline in SSH from 1971 to 2000 was reported for Mumbai²⁵.

Table 2.

The monthly and seasonal SSH details at nine geographical regions of India.

Regional SSH	Avg ± stdev (h/month)	Max (h/month)	Min (h/month)	JFD Avg ± std	MAM Avg ± std	JJAS Avg ± std	ON Avg ± std
East Coast	187 ± 46	274	72	671 ± 40	740 ± 25	574 ± 28	360 ± 24
Chennai	212 ± 44	274	167	685 ± 41	801 ± 37	723 ± 37	336 ± 36
Bhubaneswar	169 ± 54	223	73	604 ± 66	657 ± 45	400 ± 38	367 ± 35
Machilipatnam	205 ± 47	262	142	725 ± 42	760 ± 27	600 ± 32	378 ± 34
West coast	192 ± 60	290	57	727 ± 30	712 ± 40	463 ± 44	398 ± 28
Thiruvananthapuram	173 ± 44	235	111	665 ± 32	585 ± 52	538 ± 52	292 ± 40
Goa	213 ± 76	290	80	834 ± 73	788 ± 26	460 ± 83	471 ± 70
Mumbai	189 ± 71	266	57	681 ± 34	762 ± 43	393 ± 58	432 ± 23
Northern inland	187 ± 32	275	92	502 ± 55	700 ± 33	659 ± 38	392 ± 31
Amritsar	214 ± 37	275	152	502 ± 65	742 ± 43	894 ± 74	434 ± 54
New Delhi	186 ± 38	242	131	473 ± 61	702 ± 47	658 ± 66	393 ± 35
Kolkata	162 ± 46	222	90	534 ± 58	655 ± 56	425 ± 37	347 ± 34
Central inland	204 ± 57	259	104	735 ± 28	765 ± 29	533 ± 39	416 ± 29
Bangalore	169 ± 55	235	84	652 ± 35	648 ± 56	424 ± 48	307 ± 29
Nagpur	225 ± 67	290	99	773 ± 107	853 ± 98	567 ± 81	502 ± 75
Hyderabad	218 ± 54	274	128	779 ± 52	793 ± 33	608 ± 50	437 ± 41
Deccan plateau	231 ± 73	245	87	800 ± 23	892 ± 14	567 ± 47	518 ± 18
Indore	226 ± 74	309	87	767 ± 117	876 ± 102	542 ± 35	522 ± 76
Pune	223 ± 74	306	94	810 ± 28	879 ± 32	520 ± 70	473 ± 29
Ahmedabad	245 ± 72	322	100	823 ± 22	920 ± 23	639 ± 63	559 ± 22
North East	160 ± 32	229	110	530 ± 46	465 ± 29	508 ± 28	417 ± 23
Dibrugarh	161 ± 41	229	110	580 ± 50	408 ± 26	508 ± 40	435 ± 30
Guwahati	159 ± 29	203	113	480 ± 54	522 ± 49	508 ± 32	399 ± 24
Himalayan region	167 ± 50	219	87	290 ± 46	527 ± 64	831 ± 60	353 ± 49
Arabian Sea	210 ± 41	265	140	745 ± 37	701 ± 47	681 ± 66	387 ± 37
Bay of Bengal	167 ± 67	251	75	698 ± 48	601 ± 50	359 ± 27	347 ± 30

Temporal variations of SSH as anomaly, along the western coast from 1988 to 2018 are depicted in Fig. 5b. The SSH declining trend may be attribution of an increase in aerosols, resulting from human activities, alongside other factors such as region-specific changes in cloud cover, contributing to a decline in SSH⁵⁷.

Northern inland region

Northern inland is represented by Amritsar, New Delhi, and Kolkata. Northern India experiences pre-monsoon and early monsoon showers, as well as significant precipitation during the Indian summer monsoon. The routine daily rhythm for these locations includes occasional evening thunderstorms during the pre-monsoon period. The average monthly SSH for the North Indian region was 187 ± 33 h, with maximum and minimum values reaching approximately 275 h and 92 h, respectively (Fig. 3c; Table 2). Analysis of the 3-year running mean indicates an overall decreasing trend over the period 1988–2018 (Fig. 2c), intermittent slight increments for short spans of two to three years, notably between 1995 and 1996 and 2006–2008. The annual MK trend reveals declining SSH trend in the northern inland at a rate of -13.15 h/year over the span of 1988–2018.

A decreasing SSH was observed from October to December (decline from 206 to 157 h) followed by an increasing trend from January to May (increasing from 159 h to 240 h; Fig. 3c). However, with the onset of monsoon activity over the region, a decrease in SSH was noted from June (186 ± 73 h) onwards until the active monsoon season in the month of August (153 ± 58 h). Subsequently, a gradual increase was observed from September (178 ± 51 h), with a faster recovery during October (206 ± 35 h). Within northern inland region, all stations exhibit similar seasonal behaviour patterns. The regional seasonal behaviour was depicted in contrast to other regions. The Northern Inland region, located in the monsoon influenced region and overcast condition, contributing to decreased SSH in monsoon months (JJAS). Northern inland stations depict lower SSH in post-monsoon (392 ± 31 h), followed by winter (502 ± 55 h) and maximum in pre-monsoon (700 ± 33 h). The monsoon season SSH was also closer to maximum seasonal value (659 ± 38 h, Fig. 4c). The MK trend and Sen’s slope analysis defined downward trend for annual as well as seasonal SSH (Table 3), at a significance level of ≥ 95%. The temporal anomaly of northern inland region was depicted in Fig. 5c. In this region, overall negative trend was observed with a rate of -13.15 h/year. This higher decline in SSH suggesting different anthropogenic resources were active in the surroundings⁵⁷. The duration of sunshine shows a maximum over central inland and northern inland of India. Soni et al. also reported a decline in SSH from 1970 to 1990 at the Kolkata²⁵.

Central inland region

The central and southern regions of India are represented by three stations: Nagpur, Hyderabad and Bangalore, each offering a unique perspective on the climate dynamics of the area. Figure 2d depicts the SSH across all the stations over three decades (1988–2018). The annual mean SSH was found approximately 2449 ± 77 h, with maximum and minimum values reaching around 2581 and 2257 h, respectively. Over the three decades, SSH at the central inland was found a decreasing trend; however, intermittent increments were observed during the years 1995–1998 and 2006–2010 for all central inland stations. The MK trend analysis estimates an annual dimming at the rate of − 4.71 h/year.

Monthly variation in SSH (Fig. 3d) reflects the impact of both pre-monsoon and winter seasons. The pre-monsoon exerts a stronger influence on northward located stations such as Nagpur and Hyderabad, while the winter season predominantly affects Bangalore. Across all stations along the central inland, maximum SSH observed during the pre-monsoon (791 ± 43 h) followed by winter seasons (765 ± 29 h), while minimum SSH occured in post-monsoon (416 ± 29 h; Fig. 4d). The MK trend analysis (Table 3) reveals a significant negative trend for annual, winter and pre-monsoon seasons (Sen’s slope: -4.71,-2.02, -2.49, respectively) at a significance level of 95%. For monsoon and post-monsoon, MK trend and Sen’s slope depicted insignificant negative trend during 1988–2018. Figure 5d illustrates sunshine brightening trend for the span of 1988 to 2007 and then negative trend for the span of 2008–2018. The overall negative trend was observed with a rate of -4.71 h/year in central India. Although the SSH was highest over central and northern inland of India, however different rates of dimming from 1979 to 2005 in Hyderabad²⁷. Similar to this study, a significant seasonal decline in solar radiation during the pre-monsoon and monsoon seasons and a dramatic increase in winter due to prominent climatological conditions were reported in literature^21,27.

Deccan plateau region

The Deccan plateau region of India are represented by three stations Pune, Indore, Ahmedabad; each offering a unique perspective on the climate dynamics of the area. These stations predominantly exemplify a continental climate type, strategically distributed to capture the climatic nuances of Central and Southern India. Figure 2e depicts the SSH across all the stations over three decades (1988–2018). The annual mean SSH average was found approximately 2541 ± 168 h, with maximum and minimum values reaching around 2837 and 2260 h, respectively. Figure 2e shows comparatively higher interannual variability for short span but overall trend is decreasing, which is due to regional climatological variability, particularly impact of local inter annual variability in Ahmedabad station. The MK trend analysis estimates showed annual dimming at the rate of − 3.05 h/year.

Monthly SSH variation at Deccan plateau exactly mimics the monthly SSH behaviour of central inland (Fig. 3e). Across all the stations of Deccan plateau, maximum SSH were observed during the pre-monsoon (892 ± 14 h) followed by winter (800 ± 23 h) and minimum in the post-monsoon (518 ± 18 h; Fig. 4e). MK trend analysis, as detailed in Table 3, reveals significant negative trend for annual and winter season (Sen’s slope: − 3.05, − 1.44, respectively) at a significance level of 95%. For pre-monsoon, monsoon and post-monsoon, MK trend and Sen’s slope was (− 0.33, − 171, − 0.62, respectively) found insignificant negative. The temporal variations of SSH as anomaly of Deccan plateau region was depicted in Fig. 5e. The span of 1988 to 2009 was depicting brightening and 2010–2018 shows a declining trend. The inclusive negative trend observed with a rate of − 3.05 h/year in this region. A similar decline trend was reported for Pune from 1971 to 2000³². A more pronounced decline in solar radiation in Pune from 1993 to 2022 was noted²⁸.

North Eastern region

The northeast region is represented by the stations Guwahati and Dibrugarh, both centrally located and depicting very similar trends for the span of 1998 to 2018. The annual mean SSH (Fig. 2f) was 1920 ± 64 h, with maximum and minimum values approximately 2042 h and 1757 h, respectively. This part of India exhibits comparatively lower SSH in comparison to other regions. Both stations were situated within the forested areas and vegetative cover of various kinds, contributing to this observed trend. Figure 3f is depicting the monthly variation of SSH for north eastern region, revealing significantly different behaviour compared to other regions in India. An increasing trend was observed from September (133 h ± 17) to December (198 ± 27 h) and decreasing trend from January (172 ± 38 h) to April (146 ± 39 h). Whereas, relatively stable SSH trend was found during monsoonal months (118 to 138 h), except for May (163 ± 25 h) and August (138 ± 2 h), which exhibits intermittent increments.

The seasonal analysis reveals that north eastern region of India has maximum SSH during winter (530 ± 46 h), while minimum during post-monsoon (417 ± 23 h; Fig. 4f). Annual trend analysis reveals negative (Sen’s slope: − 1.33) trend. The winter and pre-monsoon MK trend was found negative. However, important contrasting observation is noticed for monsoon and post-monsoon seasons that showed positive trend (Sen’s slope: 0.13, 0, respectively) and indicate levelling off (brightening). Temporal variations in SSH as annual anomaly along the northeast from 1988 to 2018 are depicted in Fig. 5f. The lowest magnitude of SSH in northeast India can be attributed to the cloud cover and shorter daylight hours with increasing latitude²⁵. Some studies^58–61 witnessed winter and pre-monsoon decreases in SSH, may be one of the potential causes of the observed Daily Temperature Range (DTR) decreases over NE India. The SSH and DTR are the manifestation of presence or absence of the cloud. If the sky is overcast, it will lead to increase in minimum temperature and decrease in maximum temperature, generally leading to decreased DTR and SSH. In Shillong, similar insignificant seasonal decline in SSH was observed. This variation underscores the station-dependent nature of positive/negative trends due to regional sources and meteorological factors²⁵. The SSH seasonal trend in north east India, statistically shows signs of levelling off, particularly during the monsoon and post-monsoon seasons. This trend is likely influenced by a combination of meteorological, environmental, and anthropogenic factors. Deka et al. analysed linear trend from 1971 to 2010 in the upper Brahmaputra valley (in Guwahati) and revealed a decline in mean annual solar radiation, with the highest magnitude in November, similar to this study²⁴.

Himalayan region

The Himalayan subdivision represented solely by the station Srinagar, which is likely to exhibit important characteristics typical of a mountain climate. The annual mean SSH (Fig. 2g) for this region was 2001 ± 155 h/month, with maximum and minimum values reaching approximately 2283 h and 1615 h, respectively. Goel et al. also reported similar decreasing SSH in a trend analysis study for the period of 1980–2020⁴⁰. The SSH for this location exhibits an opposite configuration compared to the rest of the regions in India (Fig. 3g). It demonstrated an increasing trend from January (89 ± 34 h) to April (168 ± 28 h), a stable trend from May (219 ± 28 h) to October (209 ± 32 h), and a decreasing trend from October (209 ± 32 h) to December (87 ± 15 h). In Srinagar, the impact of monsoon is almost negligible, and so, the highest values of SSH was observed from May (219 ± 28 h) to October (208 ± 31 h). The influence of western disturbances during winter and associated cloudiness was evident in the data, particularly in the decreasing trend of SSH from October to December, followed by an increase due to reduced impact of western disturbances thereafter until April. The restructure of monthly variation into seasonal trend was depicted conflicting behaviour with the rest of the regions except to the northern inland region (Fig. 3g).

The maximum SSH observed during monsoon (831 ± 60 h) while minimum SSH occur during the winter season (290 ± 46 h; Fig. 4g). MK and Sen’s slope analysis homogenously showed negative annual and seasonal trend at a significance level of ≥ 95% (Table 3). Temporal variations of SSH as anomaly demonstrate continuous dimming for the span of 1999–2018, except year 2006, although overall annual negative trend at a rate of − 9.47 h/per year was observed for the Himalayan region of India (Fig. 5g).

Island location of Arabian sea

The island station considered in this study is Minicoy, situated in the Arabian Sea. Despite experiencing a maritime climate, the station’s unique location leads to distinct influences from summer monsoon clouds and/or winter monsoon clouds. The average monthly SSH (Fig. 2h) was 210 ± 41 h, with maximum and minimum SSH approximately 265 and 140 h, respectively. Inclusive annual decline in SSH was found at a rate of − 5.72 h/year in Arabian Sea (Fig. 2h).

Monthly variation of island location demonstrates differences, yet the impact of oceans, local cloud development, and monsoon systems was evident (Fig. 3h). Generally, there was a decrement in SSH observed from April (244 ± 22 h), instead of May (193 ± 39 h), until June (140 ± 28 h), followed by an increase from July (157 ± 29 h) onwards until February (252 ± 15 h). The recovery period was faster between July (157 ± 29 h) and December (230 ± 25 h), while it was slower between December (262 ± 16 h) and March (265 ± 15 h). These patterns reflect the complex interplay of regional climatic influences on SSH in island environment. The seasonal trend of SSH over Arabian Sea was maximum in winter (745 ± 37 h) and pre-monsoon (701 ± 47 h) as well as the lowest SSH was observed in post-monsoon (387 ± 37 h), which is analogous to all five regions of India except northern inland and Himalayan region (Fig. 4h). MK trend and Sen’s slope analysis homogenously showed negative annual and seasonal trend at a significance level of ≥ 95% (Table 3). Temporal variations of SSH as anomaly along the Arabian Sea showed that span of 1988 to 2001 depicting steady positive trend and steady negative for span of 2010 to 2018 (Fig. 5h). The overall annual MK trend was depicting negative trend at a rate of − 5.72 h/year in Arabian Sea.

Inland location of Bay of Bengal

Another island station considered in this study is Port Blair, located in the Bay of Bengal. The average monthly SSH (Fig. 3i) was 167 ± 67 h, with maximum and minimum values reaching approximately 251 h and 75 h, respectively. The influence of monsoon clouds from both southwest as well as northeast monsoons, along with numerous cyclonic activities, plays a significant role in cloudiness over the region. The months from January (242 ± 24 h) to March (246 ± 18 h) exhibit the highest absolute magnitude of SSH, followed by November (190 ± 23 h) and December (205 ± 25 h). A decreasing trend was observed from March (246 ± 18 h) to June (75 ± 17 h), followed by a slow increasing pattern from July (83 ± 22 h) to November (190 ± 23 h). Similar to the Arabian Sea, the effect of monsoonal clouds in the Bay of Bengal was more pronounced. The SSH decreased from March to June/July then increased from October (157 ± 20 h) onwards to regain higher values during January (242 ± 24 h) to March (246 ± 18 h, Fig. 3i). The seasonal trend of SSH over Bay of Bengal was depicting highest magnitude in winter (698 ± 48 h) and pre-monsoon (601 ± 50 h) and nearly similar and stable SSH observed in monsoon (359 ± 27 h) to post-monsoon (347 ± 30 h), which is analogous to Arabian Sea (Fig. 4i).

The Bay of Bengal island location also showed a typical seasonal pattern of India. The MK and Sen’s slope analysis homogeneously showed negative annual and seasonal trend at a significance level of ≥ 95% (Table 3). The inclusive annual decline occurred at a rate of − 6.10 h/year in Bay of Bengal Island (Fig. 2i).

Summary and conclusions

The Fig. 6 depicts the seasonal annual Sen’s slope trend across twenty different stations across India. Annual SSH trend across India depicts maximum decline in SSH, observed in northern inland region, particularly over Amritsar, Kolkata, as well as in the Himalayan region and West coast, specifically at Mumbai. However, clear seasonal variation was found in SSH. In winter season, whole north inland, east and west coast shows significant decline in SSH trend. Whereas, comparatively less but smaller negative trends were found in central inland, Deccan plateau, Himalayan region and north east region. Pre-monsoon season also shows overall negative trend, with a few stations, namely Bangalore, Kolkata, Mumbai, showed comparatively higher declining rates of SSH. In, monsoon and post-monsoon seasons, northeastern station of Dibrugarh shows a positive trend, in contrast to all other regions.

Fig. 6.

Annual and seasonal Sen’s slope at twenty stations across India.

This range of variation in SSH is most likely attribution of the spatial-temporal scale. Particularly, monthly and seasonal variations are directly affected by the local monsoon system and by anthropogenic and natural aerosol emissions⁶². Spatial differences in SSH across India result from varying aerosol sources, topography, and meteorological conditions. Urban and industrialized regions, such as northern India, experience pronounced decrease in SSH due to dense aerosol pollution³⁹. In contrast, coastal areas witness less SSH due to land and sea breeze driven aerosol dispersion. Monsoonal patterns influence SSH, as wet deposition during the monsoon effectively removes aerosols in monsoon regions. The north eastern region characterized by the high humidity throughout the year, while the western region generally experiences arid conditions. By closely examining SSH trends, valuable insights for better understanding of regional fluctuations can be gained. Based on the spatio-temporal trend analysis of SSH, following conclusions are drawn:

• This study examines the varying rates of decline in SSH extent over the Indian region from 1988 to 2018. The negative trends in SSH reflect the influence of local geography and climatology. The west coast, central inland, and Deccan plateau regions exhibit the highest mean SSH, with clear decline at rates of − 8.62 h/year, − 4.71 h/year and − 3.05 h/year, respectively. Conversely, the east coast and northern India showed decreasing trends at rates of − 4.88 h/year and − 13.15 h/year, respectively. The north eastern region of India experiences a decrease at a rate of − 1.33 h/year, while Srinagar, the mountainous region, showed an overall dimming rate of − 9.47 h/year. Additionally, two island locations in the Arabian Sea and Bay of Bengal also exhibit decrease in SSH at the rates of − 5.72 h/year and − 6.10 h/year, respectively.

• Our findings indicate that the east coast, west coast, and central inland stations exhibit the highest monthly SSH, while the northeastern region, Deccan plateau and Bay of Bengal record the lowest SSH due to regional weather variables.

• The monthly variation shows a typical increase in SSH from October to May with slight differences in magnitude across the regions. Subsequently, there was a significant drop in SSH noticed from June to July, followed by a slight levelling off from August to September. Contrasting trends were observed in northern inland and the Himalayan region. The northern inland region, located in the monsoon influenced region and overcast condition, contributing to decreased SSH in monsoon months (JJAS). Whereas, in contrast to all other regions, the higher SSH in the Himalayan region is due to the negligible monsoon effect, consequently increased SSH observed in monsoon and post-monsoon months.

• Seasonal negative trends were observed across all the regions, with a few exception noted during the monsoon and post-monsoon seasons in the northeast region, where a seasonal trend is positive suggesting levelling off of SSH. This trend is likely influenced by a combination of meteorological, environmental, and anthropogenic factors.

• The intra-annual anomaly plots indicate that the highest magnitude of SSH occurred predominantly before the year 2005. A steady and consistent negative anomaly, with intermittent positive anomalies lasting one or two years, was observed in all regions, except north east and Himalaya. A significant conclusion is that the magnitude of SSH decreasedeven during positive anomaly periods.

This research underscores the critical role of weather conditions in assessing solar energy potential. Moving forward, there is a pressing need for additional studies to analyse seasonal variations in cloud cover and the impact of aerosols. The findings from this study can provide valuable guidance for solar energy developers, helping them to formulate effective strategies for addressing future challenges. Solar energy developers can prioritize regions for solar panel installations based on their influence on solar radiation. Integrating these insights into planning processes can optimize renewable energy production.

Author contributions

Arti Choudhary: Conceptualization, Data interpretation, Visualization, Original draft- writing Editing and Reviewing; Bharat Ji Mehrotra: Data interpretation, Visualization; Atul K. Srivastava: Editing and Reviewing; Pradeep Kumar: Visualization, Editing and Reviewing; V.K. Soni: Editing and Reviewing, Manoj K. Srivastava: Conceptualization, Visualization, Design of experiment. All authors have read and agreed to the published version of the manuscript.

Data availability

The raw data is obtained from National Climatic Data Centre (NCDC). The datasets developed for the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.