My Research

The trend in Normalized Difference Vegetation Index (NDVI) based on the Getis-or-Gi hotspot algorithm. The figure shows Vegetation Dynamics Hotspots (coldspot as browning and hot-spot as greening). The trend in Fraction of Photosynthetically Active Radiation (FPAR) gives the photosynthetic trend (negative values for browning shown in red, positive values for greening shown in green and the white space represents the non–vegetated land for the period 2000–2019). 

The western and central India show greening, whereas the Himalaya, and the eastern and southern India show browning hotspots. The greening is found in the northwest (e.g. Punjab, Haryana), west (e.g. Rajasthan, Gujrat and Maharashtra), central (e.g. MP, regions of IGP, UP and Bihar) and east (e.g. Orissa) India. Browning is observed in the Himalaya, the eastern (e.g. Jharkhand) and southern (e.g. Tamil Nadu and Andhra Pradesh) India .

The spatio–temporal relationship of Photosynthetic activity (FPARmean) with key climate drivers namely precipitation (mm/day), temperature (C) and soil moisture content (kg/m2 ) for greening region (GR) and browning regions (BR). Here, BR–FPAR: Photosynthetic activity in browning regions, GR–FPAR: Photosynthetic activity in greening regions, BR–PRE: Precipitation in browning regions, GR–PRE: Precipitation in greening regions, BR–TEMP: Temperature in browning regions, GR–TEMP: Temperature in greening regions, BR–SM: Soil moisture in browning regions, GR–SM: Soil moisture in greening regions for the period 2000–2019. 

The years 2003–2008, 2010, 2011, 2013 and 2014 with enhanced water availability in terms of precipitation and soil moisture show greening. However, reduced precipitation and soil moisture in years 2000–2002, 2009, 2012 and 2015 exhibit browning. There is a consistent warming that influence the photosynthetic activity, and the years 2002, 2004, 2006, 2009 and 2016 experience relatively higher warming; leading to browning. The reduced precipitation together with warming have significantly reduced the soil moisture and have adversely affected the photosynthetic activity in these years. This is particularly severe in the years 2002 and 2009. Also, the years with little warming such as 2003, 2008 and 2011 show greening. The cumulative influence of enhanced precipitation and cooling cause improved soil moisture, which promote greening in these years. In sum, the changes in these drivers control the photosynthetic activity in India .

Gross Primary Productivity (GPP) trend (decreasing– browning; increasing – greening). The ratio of NPP to GPP gives the Carbon Use Efficiency (CUE) for the period 2000–2019 

Vegetation greening has a significant role in mitigation of global warming and climate change, because the terrestrial vegetation acts as a major carbon sink . South Asia has been greening in the last two decades and much of it is contributed by India and China. However, quantification of this greening in terms of terrestrial carbon sequestration is largely unknown. Carbon use efficiency (CUE) is a measure of the ability of vegetation to sequester atmospheric carbon and is estimated as the ratio of Net primary productivity (NPP) to Gross primary productivity (GPP) . It is key to understand the carbon sink and source pathways of ecosystems. It gives the amount of carbon stored and used for growth out of the net carbon acquired by the ecosystem. CUE provides insight on the vegetation functioning as it is the rate of conversion of GPP to NPP or the splitting of GPP to NPP and autotrophic respiration.

Positive trend in GPP (greening) is estimated in IGP (Punjab, Haryana and UP) northwest (Rajasthan), west (Maharashtra), CI (MP) and SI (Karnataka). Negative trend in GPP (browning) is observed in IGP (Bihar, Jharkhand, West Bengal), CI (Chhattisgarh and Orissa), NE and SI (Tamil Nadu and Andhra Pradesh). The CUE variability in India in the last two decades (2000–2019) shows a specific pattern. For instance, the regions such as NW, IGP and some areas in CI exhibit small (<0.3) CUE, but higher CUE (0.45–0.6) in CI and SI. Croplands in the western areas of SI (0.6–0.75), and forests in the Himalaya and NE (0.6–0.9) show very high CUE. Therefore, greening is found in the regions of lower CUE and browning in higher CUE regions. This is a major concern as there is a need of proper planning and management to protect the green cover in these areas of higher CUE. 

Interannual variability in Carbon Use Efficiency (CUE) and its drivers– Fraction of Photosynthetically Active Radiation (FPAR), Precipitation (P), Temperature (T), Soil Moisture (SM) and Air Organic Carbon Content (AOCC) (top panel). The Principal Component Analysis (PCA) comprising of Unrotated PCA (UPCA) and Rotated PCA (RPCA) shows the links among CUE and its drivers. Here, the green and magenta lines represent positive and negative correlations, respectively. The thickness of the lines is a measure of the strength of the link represented in terms of correlation coefficients 

The years 2003–2008, 2010, 2011, 2013 and 2014 with higher water availability (P and SM) and higher FPAR show higher CUE. However, CUE is lower in the years 2000–2002, 2009, 2012 and 2015 due to the reduction in water availability. The years of predominant warming such as 2002, 2009 and 2016 show the combined effect of limited P and high T, where the resulting low SM and  FPAR lead to small CUE (warming induced moisture stress). In addition, these years also have higher AOCC, which negatively affect CUE. The water availability accompanied by cooling has led to higher CUE (moisture induced greening) in 2004, 2011 and 2013 .

CUE (0.88), P (0.82) and SM (0.81) have a very strong positive correlation with PC1, which indicates that these variables are highly connected among themselves. T (− 0.41) has a negative correlation with PC1 as higher T will adversely affect CUE. PC2 is also distinguished by a strong positive correlation with FPAR (0.85), AOCC (0.84) and T (0.8). However, T (− 0.41) also influences PC1 negatively as higher T will adversely affect SM, P and CUE. AOCC (− 0.15) exhibits weak negative correlation with PC1, which suggests that it has a weak negative connection with CUE. Henceforth, PCA analysis demonstrates that both SM and P have strong positive association with CUE. This points out the close connection between carbon and water cycles in cropland dominated India. 

Relative influence (negative/positive) of the spatial trend in climate drivers– precipitation (drying/showering), temperature (cooling/warming), soil moisture (drying/moistening), and air organic carbon content (high/low) on the productivity trend (browning/greening) and Carbon Use Efficiency (CUE) over the period 2000–2019. 

CUE and VCD are regulated by the changes in terrestrial productivity, which is influenced by the spatial and temporal variability in drivers. SM (32.83%) is the most dominant driver of CUE in cropland dominated India; indicating a close link between carbon and water cycles. P (26.13%), FPAR (22.32%), and AOCC (16.47%) make significant contributions to CUE variability. The positive influence of SM and T dominates over the negative influence of P and AOCC; leading to increase in productivity (greening) called moisture induced greening, which is observed in NW. In CI, positive influence of AOCC might be the reason for greening due the cooling effect there, which is also replicated by anthropogenic aerosol (brown haze) in this region. The enhanced productivity (greening) in IGP can be attributed to the anthropogenic intrusions such as the improvement in irrigation facilities, enhanced farm mechanisation and application of nitrogen-based fertilizers. The negative influence of SM and T (warming induced moisture stress) is dominant over the positive influence of P, and that lead to reduced productivity (browning) in SI. In NE and eastern areas of IGP, large-scale anthropogenic activity (shifting cultivation and land abandoning) has led to green cover loss and browning. The NE region has severe consequences of human induced LULCC as the loss of vegetation cover drives extreme events such as fires and landslides in these ecologically fragile regions. 

Causal graphs representing the links for relation of Carbon Use Efficiency (CUE) with its drivers, and interrelationship among Fraction of Photosynthetically Active Radiation (FPAR), Precipitation (P), Temperature (T) Soil Moisture (SM) and Air Organic Carbon Content (AOCC). Here, the maximum time delay is Tmax and significance threshold is ά. 

P does not have a direct causal link with CUE, but it drives SM that has a direct link with CUE with lag of 2 months. P also has a link with FPAR, which has a direct positive connection with CUE in a 1-month lag. FPAR is used as a proxy for photosynthesis and higher P would support more photosynthesis and faster carbon uptake (i. e. higher CUE). P has a negative link with T (lag = 1 month), which has direct negative link with CUE (lag = 2 months). Therefore, P affects CUE indirectly through other drivers. T has a strong negative link with FPAR, as photosynthetic activity responds well only to the optimum range of T, and therefore, CUE has a direct negative link with T. Furthermore, T has a negative link with SM, which has a direct positive link with CUE; explaining the negative link between T and CUE. AOCC also has a direct negative link with CUE, as biomass burning in extensive croplands and wild fires in forests release the stored carbon stocks captured by the vegetation and reduces CUE. Similar causal links are also observed for ά = 0.05 and for ά = 0.01. 

Enhanced Vegetation Index (EVI) and Solar Induced Fluorescence (SiF) (kW m− 2 μm− 1 sr− 1 ) for vegetated land (upper panel), croplands (middle panel) and forests (lower panel) averaged over March–September in 2000–2021. Here, LD stands for the data averaged over the period March–September 2020 and PostLD stands for the data averaged over the same months, but for 2021. 

There is a steep rise in both EVI (11.54%) and SiF (16.24%) during LD (2020)as compared to that in the previous period (2001–2019). This increase in EVI and SiF is higher for croplands than that for forests during LD (2020), as the increase in EVI for forests is 5.2%, but 13.6% for croplands. Similarly, the increase in SiF (24.15%) in croplands is about four times that of forests (6.5%). Since there is an increasing trend in both EVI and SiF in recent years, we consider the period 2017–2019 as the pre-LD base-time for our analyses. 

Top panel. Spatial extent of the vegetated land in terms of croplands and forests in India, change in vegetation cover represented by Enhanced Vegetation Index (EVI), and Photosynthetic activity and productivity represented by Solar Induced Fluorescence (SiF) during LD (2020) from that in pre-LD (2017–2019). Bottom panel. Percentage (%) change in EVI, SiF, Aerosol Optical Depth (AOD) and Particulate Matter (PM 2.5) during LD (2020) from that in the pre-LD (2017–2019) for the net vegetated land (over all represents both croplands and forests), croplands and forests. 

The Indian landmass comprises of both natural vegetation such as forests and croplands. Most of its land are croplands, but forests are found in Himalaya, NEI, CI and Western and Eastern Ghats. The changes (increase/decrease) in both surface greenness (EVI) and photosynthetic activity (SiF) during LD from pre-LD. The reduction in AOD over croplands (21.11%) is higher than that over forests (15.18%) as the latter regions are remote and less polluted. Similarly, the decrease in PM 2.5 over croplands (22.53%) is also relatively higher than that over forests (21.72%). EVI (10.45%) and SiF (11%) have increased in the Indian vegetated region; consistent with the large reduction in AOD (19.27%) and PM 2.5 (22.93%) during LD. In terms of changes in forests and croplands, the increase in EVI is almost twice for croplands (14.45%) that of forests (7.91%). This difference is more pronounced with SiF, as the enhancement in SiF for croplands (17.7%) is four times that of forests (4.42%). 

Change in photosynthetic activity and productivity represented in terms of Solar Induced Fluorescence (SiF) from March to September during the three focal periods: PreLD (2017–2019), LD (2020) and PostLD (2021) to show (i) change in phonology: (a) early onset of vegetative season, (b) late withdrawal of vegetative season; (ii) Intensification (enhancement in Photosynthetic activity and productivity) during: (c) phase of decline in SiF, and (d) phase of increase in SiF. The regions marked in red show the particular crop type, based on data from Ministry of Agriculture, Government of India (GOI). The red line (upper panel) represents the mean SiF value for the period 2017–2019 (baseline). 

The enhanced surface greenness and photosynthetic activity during LD is due to the extended growing period (phenology). the first event marked as “a” shows that the start of the season (SOS) during LD is 8 days earlier than that in pre-LD in most of India. The SOS is more than 16 days earlier in some regions such as Himalaya, NEI, IGP, NW and SI. The rest of the vegetated regions depict enhancement in SOS by 8–16 days. The 16 days earlier onset of growing season of wheat in Punjab, Haryana and North Bihar is crucial. The NW region has mixed crops such as food crops (maize and soyabean) and cash crops (cotton), which show 16 days earlier SOS. The mixed crops (jowar, bajra and millets) in SI also show similar early onset. The next event marked as “b” shows that the end of the season (EOS) is extended by 16 days or more during LD from that in pre-LD. The EOS is prolonged by 16 days for lower Himalaya, NEI, north west IGP, western areas of CI and most regions of SI. This is particularly important for rice growing areas in Punjab and North Bihar. The rest of the vegetated land, mainly in CI, NW and eastern IGP, have EOS delayed by more than 16 days. The NW has crops, such as coarse cereals and cash crops (cotton and oilseed), with EOS extended by more than 16 days. SI has cereals and oilseed; showing similar delay in EOS. Therefore, it is clear that some croplands in IGP, NW and SI have prolonged growing season by about 32 additional days. Similarly, some regions in CI show extended growing season by about 24 days. 

The enhanced surface greenness and photosynthetic activity during LD from pre-LD are also due to intensification through higher rates of photosynthetic activity and thus, enhanced productivity. The intensification is represented in terms of higher SiF during the same days in LD from that in pre–LD. This intensification is marked at two points “c” and “d”. Here, “c” represents the intensification during the phase of declining SiF and “d” represents the intensification. during the phase of rising SiF. The photosynthetic activity has enhanced in all regions during the phases of decline and rise in SiF. During the phase of declining SiF, the areas depicting high change in photosynthetic activity are observed in most of Himalaya, some regions of IGP, NW, CI and Western Ghats. The rest of vegetated regions show very small or no increase in photosynthetic activity. During the phase of enhanced SiF, Himalaya, NEI, IGP, NW, eastern areas of CI and Western Ghats exhibit high change in photosynthetic activity. Himalaya has pine and oak in the west along with some evergreen broadleaf and needled forests in the east, which exhibit high increase in photosynthetic activity. The wet Savannah in Western Ghats also shows high increase in photosynthetic activity during both “c” and “d”. 

The COVID–19 induced LD served as natural experiment to expose the close links between air quality and surface greenness. Our analysis shows that reducing the intensity of human activity can cause rapid response in the environment. The increase in vegetation, particularly in croplands in response to the improvement in air quality is a new insight. In future, maintaining good air quality will serve not only good health, but also a sustainable environment for future generations. As a result, serious efforts must be made to monitor vegetation dynamics at various spatial and temporal scales to fully comprehend the effects of both natural and anthropogenic forcings on vegetation cover, photosynthetic activity and terrestrial productivity. These efforts will allow us to address the issues of climate change, global sustainability and food security.

Since India is very rich in vegetation with vast croplands and forests, such studies are very important. Therefore, we analyse the impact of changes in air quality on surface greenness and photosynthetic activity in both croplands and forests of India. We observe that both vegetation and photosynthetic activities have increased due to the improvement in air quality. The enhancement in croplands is more pronounced than that of forest, which is also very crucial for a country like India where half of its people are practicing different forms agriculture that contribute about 16% to its GDP (Gross Domestic Product). As there are many countries in the world with similar environmental and socio-economic conditions (e.g. high population, severe pollution and agriculture-based society), our results have global significance and wider implications. Henceforth, this study would help policy-makers to draft laws for improving air quality, not only to protect public health and to mitigate climate change, but also to enhance crop yields for regional and global food security.