1 Introduction

Forest resources are in turmoil because of climatic variations and anthropogenic actions, which compel as a high priority (Gupta & Pandey, 2021). Globally, forests are categorised as protected or trees outside forests (unprotected forests) (Leberger et al., 2020; Thomas et al., 2021). However, forests are a crucial component of the biosphere, which provides habitat for living beings and is the centre of numerous ecosystem services. They have several elements, such as tree cover, shrubs, herbs, wildlife species, tiny organisms, and the ability to regulate climatic factors (Surov & Kuelka, 2019). Even though, the protected (reserved) forest area has more tree cover, which made suitable for wildlife’s. It is important to protect and maintain ecological integrity, as urban green spaces serve as biogenetic reservoirs and are considered indicators of environmental sensitivity for sustainability at different spatial scales (Egloff, 2006; Leberger et al., 2020).

Somehow, forest health is a crucial driver of the ecosystem or landscape that regulates the regional climatic, socio-economic, and environmental conditions of the region (Saha et al., 2021; Trumbore et al., 2015). To understand the forest's response to the climate and environment and then assess its health quality. Forest health was first defined by Laughlin (1994) as "the capacity of the land for self-renewal" conservation is our effort to understand and preserve this capacity. Morelan (1994) again described forest health to understand sustainability and its meaning. Forest health is influenced by a combination of biotic and abiotic factors, including elements such as temperature, precipitation, CO2, and others, as well as factors like vegetation type, structure, canopy density, wildlife density, age, and others (Trumbore et al., 2015). However, changes in climate, specifically rainfall and temperature, have had a noteworthy impact on forest health at the regional level (Hartmann et al., 2022; Pause et al., 2016). Therefore, the current study concentrated on a protected area in Rajasthan, which is in the western part of India. It is essential to evaluate and monitor the health of the forest effectively to ensure sustainable management.

In recent advances, forest health has been assessed based on the biophysical or biochemical attributes of their vegetation types with the help of emerging technologies such as remote sensing, drones, and others (Lausch et al., 2016). Remote sensing technology provides a window to extract information without reaching a location with accuracy (Barmpoutis et al., 2020). It is possible to assess the spatiotemporal information of earth features such as forests, bodies of water, deserts, distinct ecosystems, and others (Chowdhury, 2006; Pause et al., 2016). Besides that, in forestry, remote sensing approaches or methods on different resolutions (i.e., spatial, temporal, and spectral) datasets to obtain precise results. The optical remote sensing dataset is widely used for forest health because it is more convenient and straightforward than other remote sensing datasets. So, most of the studies were carried out using change detection, spectral indices, Land Use Land Cover change and their statistical analysis (Woodall et al., 2010; Pautasso et al., 2015; Lausch et al., 2016). Also, detailed analysis integrates remote sensing with climatic parameters (Gupta & Sharma, 2019; Raj & Sharma, 2022). Remote sensing offers considerable advantages in assessing forest health across various dimensions and conservation areas. It allows for the efficient and economical surveillance of expansive forest areas, aiding in the identification of alterations in vegetation composition, biomass, and health markers such as leaf chlorophyll concentration and canopy moisture stress (Raj & Sharma, 2022). Moreover, remote sensing methods like multispectral and hyperspectral imaging, LiDAR, and thermal infrared offer crucial data on forest biodiversity, ecosystem operations, and reactions to disruptions such as wildfires and pest outbreaks (Li et al., 2021; Senf et al., 2017). Nonetheless, challenges remain, such as constraints in spatial and spectral resolutions, atmospheric disturbances, and the necessity for terrestrial verification (Zhang et al., 2005). Additionally, analysing remote sensing data requires proficiency in data processing and ecological knowledge to guarantee precise and significant evaluations of forest conditions (Dutta et al., 2020). Despite these hurdles, remote sensing continues to be an essential tool for observing and managing forest health across distinct topographical conservation areas.

Furthermore, forest health in protected areas is vulnerable due to climate change and air pollution (Smith et al., 2020; Tabor et al., 2018). So, India currently has 5.27% of the total area as a protected area, including national parks, sanctuaries, conservation reserves, and communities’ reserves. However, the current study over the 'Sariska National Park' (SNP) is one of the protected areas that is reserved for tiger habitat, especially. Earlier, studies mainly focused on endemic species, tigers, birds, herbs, and shrubs in the SNP area (Jain et al., 2009, 2016; Yadav & Gupta, 2007; Rani & Kumar, 2011). After 1990, certain studies analyzed vegetation cover classification (Sharma et al. 2013) and assessed the SNP's biomass using remote sensing techniques. In 2014, a study in SNP critically evaluated the reserved area habitable for birds and biodiversity, now encroached by human activities (De & Chauhan, 2014; Shahabuddin et al., 2004). However, several studies in the SNP areas based on fields, social, flora, and fauna species were reported in Table 1. Over the years, no such studies have been conducted that attempt to understand the forest health of SNP using a remote sensing approach.

Table 1 Important ecological studies on Sariska national park after 2000’s
Full size table

Furthermore, this study uses a geospatial approach to ascertain the spatiotemporal health of the SNP-protected area in qualitative and quantitative aspects. Insofar, we evaluate spectral vegetative indices (i.e., NDVI, DVI, MSAVI-2, and MSI) for the years 1972, 1996, and 2018 to understand spatial forest health based on reflectance values. Also, assess Land use Land Cover change to derive different classes and the change between them from 1972 to 1996, 1996 to 2018, and 1972 to 2018. Here, it monitors forest cover change (open and dense) between 1972 and 2018 and its conversion. In addition, this study introduced an integrating approach of climate parameters with vegetative indices such as NDVI and MSAVI with their coverage areas. To investigate the understory health response using integrative rainfall and MSAVI-2, as well as to understand forest greenness health responses in the region over the last 46 years using integrative rainfall and NDVI. Also, understand the climatic trends of rainfall and temperature and their effects on the SNP's health between 1972 and 2018. Hence, it helps forest managers manage forest resources and areas and is an essential indicator for protected area management and wildlife conservation.

2 Materials and methods

2.1 Study area

Sariska National Park was declared as a Tiger Reserve in 1978 by the Government of India, and it has one of the most suitable habitats for tigers in India. It is a part of the upper zone of the Aravalli hill range and is situated in the Alwar district of Rajasthan state (Jain et al., 2016) with geographical coordinates 76° 28′ E and 27° 32′ N (Fig. 1). It has a subtropical evergreen dry deciduous forest type with tree and shrub heights ranging from 2 to 25 m and a genetic pool (De & Chauhan, 2014; Dang, 2005). However, its distinct topography contains valleys and hills ranging in elevation from 550 to 770 m above mean sea level (MSL). The pedological nature of SNP is sandy-loam with alkaline (Yadav & Gupta, 2007), and it receives low annual precipitation of 65–70 cm and an annual mean temperature of 16–34 °C. The SNP total area is 1479.01 km2, based on the 1956 toposheet map of the Survey of India. There are 26 villages, 12 of which are in critical tiger habitats. Two state highways pass diagonally through the SNP, e.g., Alwar–Thanagazhi–Jaipur, and Sariska–Kalighati–Tehla, which cover 44 km of the reserve area. In 1978, Sariska Tiger Reserve was included in the list of tiger reserves by the Government of India, and subsequently, the Ministry of Environment and Forests issued guidelines to the state government to delineate the critical tiger habitat on the basis of abundant vegetation and tiger habitat suitability (Jain et al., 2016).

Fig. 1
figure 1

Study area: a India, Rajasthan and b Rajasthan with Alwar Dist. c Alwar with SNP, d DEM of Sariska National Park

Full size image

SNP is classified into two zones (the buffer zone and the core zone). A buffer zone means the periphery which comes under the protected area in which better development measures have been taken for enhancement of the conservation value of the area’ and the core zone is defined as' the regions of a protected area which contain suitable habitat for numerous flora and fauna, including higher-order predators with minimal conflict’. The critical tiger habitat was divided into three core areas: Core I (273.8 km2), Core II (126 km2), and Core III (97.5 km2) "(Semlitsch & Jensen, 2001). SNP is dominated by four floral communities: trees, shrubs, grasses, and weeds (Kidwai, 2013) while the most visible wildlife’s are Tigers, Leopards, Dholes, and Chital (Shahabuddin et al., 2004).

2.2 Datasets and image interpretation

To analyze decadal forest health assessments of SNP over the last 46 years, satellite data of LANDSAT-MSS, TM, and Sentinel-2A from 1972, 1996, and 2018 were downloaded from the USGS (Table 2) website (https://earthexplorer.usgs.gov). Accordingly, the continuous climatic data downloaded from (https://www.indiawaterportal.org/met_data/) and (https://power.larc.nasa.gov/data-access-viewer/). Landsat satellite series imageries were used for the following years: 1972, 1996, and 2018. The available images are from the months of February to May, when there is the least amount of cloud cover. Datasets that do not have the month's exact date have needed some image ratification (image ratification and radiometric correction). Preprocessing of the satellite imagery like radiometric correction (calibrate the pixel value for enhanced analysis quality), image enhancement (color conversion and intensity correction noise removal), pan-sharpening (imagery of the year 1972 with the year of 1996 to continue with an equal pixel value of 30 m) and mosaicking tiles were done using ArcGIS 10.5. In ERDAS Imagine 15.1, software was used for cloud filtering of the 2018 and 1996 imageries, then clipped the Area of Interest (AOI) SNP from preprocessed imageries. After preprocessing of the imagery, it was used to retrieve vegetation indices and evaluate LULC classification. Preprocessed images for Land Use and Land Cover (LULC) assessment using Mahalanobis Distance Classification (MDC) are based on Euclidian minimum distance with direction sensitive to band pixel, which classifies imagery based on a covariance training features matrix (Ahmed et al., 2015). Therefore, the MDC approach was used for image classification in the current study for better accuracy. Based on image MCD classification, SNP has been classified into six major classes: dense forest, open forest, settlement, agricultural land, barren land, and water bodies. However, open and dense forest class categories are based on tree cover density (FSI, 2019). As a result, more than 40% of tree cover is classified as dense forest, while less than 40% is classified as open forest. In addition, open (0.1–0.4) and dense (0.4–0.8) NDVI values were used. Additionally, for LULC accuracy, the Kappa Coefficient was obtained for three decadal periods. Also, we used high-resolution imagery from Google Earth for the year 1996, as well as ground-truthing for the year 2018; however, for the year 1972, an available report was used instead of SNP. Forest cover type and density maps were delineated using vegetative spectral indices to evaluate the forest health of the SNP. The overall paradigm for forest health in SNP between 1972 and 2018 is shown in Fig. 2.

Table 2 Details of satellite imagery
Full size table
Fig. 2
figure 2

Paradigm for forest health assessment of SNP in between 1972 and 2018

Full size image

2.3 Vegetative spectral indices

Vegetation plays an essential role in the earth's climate system. Alterations in vegetation cover could cause variations in land surface temperature, precipitation, and others. The plant’s chlorophylls absorb the visible range of the electromagnetic spectrum; however, their reflection is low in blue and green, whereas their high reflection is in red (0.58–0.68 µm) and 3–1. 1 µm. Therefore, the vegetative indices are formed by a mathematical amalgamation of visible and near-infrared bands of satellite imagery. This range of absorbance and reflectance of vegetation indicates its environmental condition, canopy cover, and growth phase change (Golubyatnikov & Denisenko, 2006). Vegetative indices are used to delineate the greenness and forest health of a region. Currently, four vegetative spectral indexes are used: NDVI (Normalized Difference Vegetation Index), is a ratio of the red band and the NIR band that shows the regions' greenness, and its value range lies between − 1 to + 1 (Rouse et al., 1973). The DVI (Difference Vegetation Index) is another vegetation index that indicates the vegetation's chlorophyll content in a region; its value range lies between 0 and 3 (Pearson & Millar, 1972). MSI (Moisture Stressed Index) is a vegetative and a hydrological index that specifies water availability (Rouse et al., 1973). The MSAVI-2 (Moist Soil Adjust Vegetation Index-2) is the vegetative index, which is an advanced form of the SAVI (Soil Adjust Vegetation Index); it indicates that the canopy intensity has a range value that lies between − 1 to + 1 (Qi et al., 1994). It can help forest managers and ecologists detect early signs of stress or degradation in forests. By combining near-infrared and red light reflectance in a specific way, MSAVI2 highlights areas of healthy vegetation while reducing the influence of soil background noise.

$${\text{NDVI }} = {\text{ NIR}} + {\text{RED}}/{\text{ NIR}} - {\text{RED}}$$
(1)
$${\text{DVI }} = {\text{ NIR}} - {\text{RED}}$$
(2)
$${\text{MSI }} = {\text{ NIR}} + {\text{SWIR}}/{\text{SWIR}} - {\text{NIR}}$$
(3)
$${\text{MSAVI }} - {2 } = \, \left( {{2 }*{\text{ NIR }} + { 1 }{-}{\text{ sqrt }}\left( {\left( {{2 }*{\text{ NIR }} + { 1}} \right){2 }{-}{ 8 }* \, \left( {{\text{NIR }} - {\text{ R}}} \right)} \right)} \right) \, /{ 2}$$
(4)

2.4 Integrative analysis of vegetative index and climatic parameter

Forest health depends on biotic and abiotic factors (Raj & Sharma, 2023; Saha et al., 2021). Depending on their geographical location, different climatic parameters like precipitation, temperature, and solar irradiance affect forest health. This study proposed an integrative approach of spectral indices and climatic parameters efficiently assesses the forest health of the region. Precipitation is the one of the leading climatic parameters that affects forest growth, and the change in temperature is significantly unable to become a limiting factor during the study period. The vegetative indices NDVI and MSAVI-2 mean values are integrated with mean precipitation for the vegetation covered area. Between 1972 and 2018, data on climatic parameters (temperature and precipitation) were used to access the climatic parameter trend and correlate temperature fluctuations to the SNP precipitation pattern. Consequently, 30 random data points were selected from the vegetative raster indices based on topography and taken its mean value and stationed data values of climatic factors were plotted accordingly of the years 1972, 1996, and 2018. Also, vegetative indices classify based on climatic factors observation such as 0 to 0.25, 0.25–0.5 and 0.5–0.75 and 0.75 to 1 and calculate areas and plot accordingly. Additionally, it signifies forest health in 1972, 1996 and 2018. NDVI and MSAVI-2 vegetative indices integrate with rainfall of a region because rainfall is a crucial parameter that limits forest growth due to topographical effects. This analysis is based on vegetation NDVI area cover, value and the annual rainfall range (low to high).

3 Result

3.1 Land use/cover of SNP

To understand the change and its health map in the SNP, first analyzed the Land Use Land Cover for the years 1972, 1996, and 2018. So, it has been classified into six classes: dense forest, open forest, water bodies, agricultural land, settlements, and barren land. In the last 46 years, they found a significant change in each land class. The proportion of settlement, agricultural land, and dense forest has changed the most. The land use land cover of SNP evaluated in 1972, 1996, and 2018 is shown in Fig. 3. In 1972, the dense forest class accounted for 14%, open forest 73%, water bodies 5.53%, agricultural land 1.02%, and a combination of settlement and barren land rounded out 5.84% of the total protected area. In 1996, it was observed that dense forest was 11.80%, the open forest was 70.30%, water bodies were 8.15%, the agricultural land area was 3.08%, and the combination of settlement and barren land was 6.46% of the total protected area. While in 2018, it was observed that open forest was 54.96%, the dense forest was 32.06%, water bodies were 0.7%, agricultural land was 4.08%, and the combination of settlement and barren land was 4.66% of the total protected area. LULC assessment, on the other hand, requires an accuracy assessment and validation. Table 3 shows the land use and cover in terms of area for the SNP years 1972, 1996, and 2018, as well as the accuracy assessment.

Fig. 3
figure 3

LULC map of SNP year 1972, 1996, 2018 with their classes

Full size image
Table 3 LULC classes areas (Km2) wise of the year 1972, 1996 and 2018 with accuracy assessment
Full size table

3.2 Forest health status

Forest type change is a crucial indicator for forest health, delineating the current forest cover type status. This protected area has changed from open to dense, and dense to open forest evaluation is based on the land use/cover classification of 1972 and 2018. Between 1972 and 2018, the dense to open forest conversion area is shown in "red," the open forest to dense forest conversion area is shown in "green," and the unconverted areas are shown in "black. "The open forest has become densely forested in these critical areas, except for the central part; the dense forest has become an open forest. According to Fig. 4b, open to dense forest-converted areas were 453.33 km2 (38.15%); dense to open forest-converted areas were 793.90 km2 (53.67%), and areas that remained unchanged in the last 46 years were 121.87 km2 (8.18%). The growth of dense forests in the SNP has created a better environment for wildlife over the last 46 years, as forest health has improved. Dense forest cover increased in 2018 due to a rebound project that started with the Asian Development Bank (ADB) and the Rajasthan government in 1998–2002.

Fig. 4
figure 4

a Change detection map between 1972 and 2018 of SNP with, b their change area bar-graph

Full size image

3.3 Climatic variables analysis

Climate acts as an abiotic factor influencing forest growth and regulation of an ecosystem. So, the current study used rainfall and temperature, both of which are important contributors to forest health in any geography. Thus, topography and climatic factors in this protected area contribute to maintaining ecological diversity and forest health. Figure 5 depicts annual rainfall and temperature data from 1972 to 2018 and their trend lines.

Fig. 5
figure 5

Climatic variables (Rainfall and Temperature) of SNP from 1972 to 2018

Full size image

However, changes in rainfall patterns have an impact on vegetation and are responsible for a region's seasonal temperature. The annual average rainfall pattern gradually changed between 1972 and 2018, but the lowest observed rainfall was 27 cm in 1986 and 2003, and the highest rainfall was 99 cm in 1974. While the observed rainfall trend line, the intensity decreased from 2010 to 2018, indicating that it was a responsible factor in SNP forest health after 2010. Another variable temperature has remained stable in this protected area, with an average annual temperature ranging from 24–30 °C. As a result, the temperature regime within the protected region does not appear to exert a significant influence on the critical forest health status of the SNP in the last 44 years.

3.4 Vegetative indices

This study used four different types of spectral indices, which give adequate information about vegetation, i.e., NDVI, DVI, MSAVI, and MSI; whereas MSI is a hydrological index, it is used to determine water availability in forest areas. Vegetation indices are rationing bands that sense ground reflectance values of the vegetated area. From Fig. 6, the observed NDVI maximum range value in 1972, a range value between 0.07 and 0.58, determined the maximum open forest type areas in SNP. In 1996, its range value between 0.07 and 0.8 had shown healthy successive growth in 24 years. In 2018, values of 0.18 to 1.0 and 0.52 indicated the area has a moderate vegetation type. MSAVI-2 provides essential knowledge about forest health. From Fig. 7, it was observed that in 1972, its value between 0.4 and 0.8 showed less canopy density, which meant the health of the forest wasn’t efficient. But in 1996, the range value was between 0.2 and 0.7 and represented an increase in the vegetation canopy density over the region's health for 24 years has improved. In 2018, the value range mostly lay between 0.1 and 0.5; The canopy density increased after 1996 due to the high vegetation intensity, which was not observed in the soil in 2018. The forest health dynamics moved toward greenery in 46 years and made it viable to the region and wildlife. From Fig. 8, it was observed that in 1972, the DVI range had mostly fallen to 1.0–2.2 notified vegetation, which was present but not broadleaf or dense in nature. In 1996, the value ranged between 0.7 and 2.7 and showed more forest area than the previous record. In core zones, broadleaf vegetation increased after 2018. The value range between 0.8 and 3.0 represented that mostly vegetation was healthy in the region, and their growth was in a significant direction. From 1972 to 2018, the DVI value gradually rose and constructively displayed the forest's dynamic growth. From Fig. 9, the observed MSI value range in 1996 was between 0.1 and 1.8, showing the region had a low water stress region. But in 2018, the range value lies between 0.1 and 2.8, representing a slightly high-water stress region due to deficient precipitation. In 1996, annual precipitation was high, and in 2018, the precipitation was low. The comparison showed that forest health in 1996 was better than in 2018, depending upon yearly precipitation validated by the MSI index value range. However, the 1972 MSI map was not emulated because Landsat—MSS does not have SWIR band which is essential for MSI.

Fig. 6
figure 6

Illustration of NDVI map of SNP year 1972, 1996 and 2018

Full size image
Fig. 7
figure 7

Illustration of MSAVI2 map of SNP of year 1972, 1996 and 2018

Full size image
Fig. 8
figure 8

Illustration of DVI map of SNP of year 1972, 1996 and 2018

Full size image
Fig. 9
figure 9

Illustration of MSI map of 1996 and 2018 (1972 data not available)

Full size image

3.5 Vegetative indices with climatic variables

The intricate relationship between rainfall, NDVI, and MSAVI2 indices is of paramount importance in deciphering forest health dynamics. Climatic (abiotic) is crucial for forest health dynamics whereas rainfall is most important in them, which directly affects the growth of forest. As rainfall abundance increases, vegetation tends to flourish, resulting in increasing NDVI values attributed to heightened near-infrared reflectance from robust vegetation cover. This positive correlation indicates healthy forest growth. However, the relationship between rainfall and MSAVI2 is more complex. MSAVI2 is designed to minimize the impact of bare soil on the index value, making it useful for forests with exposed soil. During drought conditions or low rainfall periods, NDVI values decline due to water stress, while MSAVI2 values can provide insights into the extent of vegetation stress and potential forest degradation. This complementary use of NDVI and MSAVI2 allows us to assess forest health comprehensively and detect early signs of forest degradation.

3.5.1 NDVI and rainfall

Based on the integrative approach Fig. 10 graph shows a potential relationship between rainfall, NDVI, and the area. NDVI indicates the vegetative area, whereas rainfall is a climatic factor that helps grow vegetation of a region. In 1972, from Fig. 10, maximum areas covered under NDVI range value 0.02–0.25 whereas rainfall between 20 and 50 cm, while in 1996, maximum area covered under NDVI range value 0.02–0.50, having rainfall 65–110 cm. Lastly, in 2018, maximum areas covered under NDVI ranged from 0.01 to 0.80, and rainfall ranged from 40 to 70 cm.

Fig. 10
figure 10

Integrated graph of Rainfall and NDVI of the SNP of year 1972, 1996 and 2018

Full size image

3.5.2 MSAVI2 and rainfall

Based on the integrative approach Fig. 10 graph shows a potential relationship between rainfall, MSAVI2, and the area. MSAVI2 indicates the vegetation health and vigor of a region. In 1972, Fig. 11 observed that maximum areas covered under 0.02–0.12, whereas rainfall was 10–30 cm, while in 1996, maximum areas covered under the MSAVI-2 value range was between 0.1 and 0.62, having rainfall between 70 and 120 cm. In 2018, it was observed that maximum areas were covered under a range value between 0.01 and 0.4, whereas rainfall was 30–60 cm.

Fig. 11
figure 11

Integrated graph of Rainfall and MSAVI2 of the SNP of year 1972, 1996 and 2018

Full size image

4 Discussion

4.1 Current study and comparative analysis

The current study examined the spatial–temporal forest health status of Sariska National Park, which is located in the upper Aravalli range, in 1972, 1996, and 2018, using multispectral remote sensing datasets fused with climatic variables. Some studies of remote sensing indices have demonstrated different degrees of sensitivity to leaf content reflection variation under diverse scenarios and canopy types (Lechner et al., 2020). However, several studies on forest health are based on different regional levels, such as national parks, sanctuaries, ecosystems, landscapes, and biomes. In addition, these studies were conducted in a distinct region of India on eco-regions, such as national parks, wildlife sanctuaries, and bio-reserves. Therefore, from Sect. 4.1, the current study analyses the land use land cover change of SNP between 1972 and 2018, revealing that in their classes, such as settlement, dense forest, and agricultural land, areas increased by 51.34, 261.3, and 45.24 km2, respectively. While water bodies, barren land, and open forests decreased by 77.5, 19.14, and 267 km2, respectively. Similarly, in the study by Tiwari et al. (1990), 51.82% of the area was densely forested, and 28% of the area was covered with open vegetation, representing 53–55% of the total area from 1989 to 2014 was forestland of SNP. Also, Sharma et al. (2016) used a forest fragmentation approach to study Khangchendzonga Biosphere Reserve in Sikkim between 2000 and 2019 and found that land cover dynamics show a decrease in open forest, alpine scrub, alpine meadows, snow, and hill shadow areas by 2.81, 0.39, 8.18, 3.46, and 0.60%, respectively, while an increase in dense forest and glacier area by 4.79 and 10.65%, respectively. In addition, Ramachandran et al. (2018) analysed the LULC of a protected area of the Western Ghats, that is, Kudremukh National Park (KNP) and Bandipur Tiger Reserve (BRT), between 1973 and 2016. Thus, in the KNP region, evergreen forest cover increased during 1973–2016 from 33.46 to 27.22%, whereas in the BRT, deciduous cover increased from 61.69 to 47.3% due to mining, horticulture plantations, human habitations, etc. Talukdar et al. (2019) study carried out using a forest fragmentation approach on the Patharia Hills Reserve Forest of northeast India between 1988 and 2016 reported that 10.52% of the forest cover area was increased overall.

Moreover, in Sect. 3.2 analyze, forest health by spectral indices i.e., NDVI, DVI, MSAVI-2, and MSI for the years 1972, 1996, and 2018. In the SNP region, the 1972 average value was 0.325, but in 2018, the average value was 0.59, indicating that SNP health improved significantly. Meanwhile, MSAVI-2 determines soil-exposed areas with the status of sparse vegetation in the region; in 1972, its mean value was 0.6, whereas in 2018, its mean value was 0.3, which indicates a considerable change in vegetation. DVI is sensitive to the spatial amount of vegetation; it is not distinct which type of undergrowth; in 1972, its mean value was 1.6, while in 2018, it was 1.9, which reveals overall vegetation is in growth. Finally, the MSI-derived available water content of the region in 1972 was 0.95, and in 1996 it was 1.45, indicating adequate water availability. Similarly, Saha et al. (2021) analyzes forest health based on spectral indices such as NDVI, NDMI (Normalized difference infrared index), EVI (Enhanced Vegetation Index), SAVI (Soil-adjusted vegetation index), SI (Shadow Index), BI (Bareness Index), and GI (Greenness Index) of Buxa reserve forest, sub-Himalayan region. It revealed that forest health based on indices of years 2001 and 2019, the BRF (Buxa Reserve Forest) drastically changed due to human intervention. In 2019 only a few patches of high-elevation areas were covered with greenery and dense vegetation. Another study performed by Ahmad et al. (2019) on the Sholayar Reserve Forest of Kerala state using hyperspectral datasets derived indices such as MSI, EVI (Enhanced vegetation index), NDNI (Normalized difference nitro Index), WBI (Water band index), ARVI (Atmospheric resistant vegetation index), ARI1 (Anthocyanin reflectance index 1), CRI1 (Cellulose absorption index), RENDVI (Red-Edge normalized difference vegetation index) and VREI-1 (Vogelmann red edge index-1). Based on these indices, it was concluded that around 218.33 km2 of the study area is under the category of a moderately healthy forest, 141.53 km2 is under good healthy condition, and 77.2 km2 is found to be in the least healthy state. Dutta et al. (2017) carried out a study over the Durgapur Forest range, West Bengal using land use land change (fragmentation approach) and NDVI for 25 years and reported that open forests remarkably decreased, and dense forest patches were significantly affected by anthropogenic intervention. In another study by Gupta and Pandey (2021) forest health was analyzed based on spectral indices i.e., ARI1 (Anthocyanin reflectance index), Structure insensitive pigment index (SIPI), Canopy Chlorophyll Content (CCC), and Normalized difference vegetation index (NDVI) over Chotanagpur terrain region, Jharkhand. It has been reported that forest health increases by 6 and 3% in the winter season compared with the summer season. Mahato et al. (2021) performed studies over Simlipal National Park, Odisha, based on vegetative indices such as NDVI, MSI, Modified Chlorophyll Absorption Ratio (MCARI), and Soil and Atmospherically resistant vegetation index (SARVI), which provide maximum accuracy for identifying vegetation classes, with the eastern and central parts of the study area having excellent vegetation cover.

Furthermore, Sect. 4.3 established the relationship, based on an integrative approach, between the spectral indices and climatic parameters that elicit forest health status and vegetation growth affected by climatic variables over the last 46 years. So, we discussed the interdependence between NDVI & rainfall and MSAVI-2 & rainfall. According to Fig. 9, NDVI values between 0.3 and 0.5 covered the majority of the area where rainfall was deficient between 20 and 50 cm in 1972. In 1996, the rainfall value range increased to 120 cm, leading to areas with sharp NDVI values. In 2018, the rainfall value range increased to 70 cm, covering most areas with 0.3–0.6 NDVI values. We also discuss the continuous rainfall and temperature profiles of SNP from 1972 to 2018 to understand the relationship between rainfall and temperature and its dependence on vegetation. Again, in Sect. 4.4, Fig. 10 shows that the rainfall trend prior to 2010 followed a random pattern but was higher than the average. However, temperature is one of the sensitive parameters of forest health. As a result, 1.2 °C has risen in the last 46 years, demonstrating the effects of climate change and anthropogenic action on SNP vegetation. In Sect. 3.5 analyzed the interrelation between the vegetative indices and rainfall. This provides an intricate relationship between NDVI and rainfall, in which 1972 the 2018 trend of forest greenery with respect to rainfall increase between 1972 and 1996 and then slightly decrease in 1996–2018. However, MSAVI2 and rainfall describe forest heath stress and degradation. Therefore, in last the 44 years forest health has been under stress due to decrease the rainfall between 1996 and 2018. Hence, overall, the forest health of SNP has improved. Although forest health has improved due to a positive trend in rainfall, the government has taken the initiative to rebuild the green wall of Aravalli in the 2000s, either around or within tiger-reserved areas. Several NGOs and other organisations have implemented and managed continuous ground monitoring strategies.

4.2 Advancement and limitation of this study

The advancement of studies on forest health and land-use land-cover change (LULC) in protected areas, such as Sariska National Park (SNP), is crucial for understanding ecosystem dynamics and tiger and other endemic species conservation. This current study builds upon previous research by examining the spatial and temporal changes in SNP's forest health and LULC over three distinct periods: 1972, 1996, and 2018. The integration of multispectral remote sensing datasets with climatic variables offers a comprehensive perspective. The study further inquired into the intricate relationship between spectral indices (NDVI and MSAVI-2) and climatic variables, specifically rainfall and temperature. It was observed that NDVI values corresponded to rainfall patterns, with higher values associated with increased rainfall. This relationship provides insights into vegetation health and growth. Additionally, the analysis of temperature trends revealed a concerning rise of 1.2 °C over the last 46 years, indicative of climate change and anthropogenic influences.

Forest health evaluations require continuous monitoring and conservation strategies to ensure a sustainable environment and ecology. This study evaluated the forest health status of Sariska National Park (SNP) using remotely sensed multispectral datasets and integrated them with climatic parameters to assess them accurately and determine their interdependence. In this study, we used optical platform (Landsat series and Sentinel-2) datasets and derived LULC and vegetation health indices (NDVI, DVI, MSAVI, and MSI) to determine the forest health of SNP between 1972 and 2018. However, the current study did not include NPP or biomass, which reflect growth health and other indices. The previous section discussed previous remotely sensed and ground studies related to the ecological health of the SNP. Therefore, this study was conducted both quantitatively and qualitatively.

4.3 Scope of current study

Furthermore, future studies need to focus on low temporal scales and other high spectral and spatial resolution datasets, such as Hyperspectral and LiDAR. In addition, for biomass, we used microwave datasets with low temporal and high spatial resolutions in the geospatial approach. Future scientific communities will need to adopt state-of-the-art technology and modelling approaches for proper management planning and conservation. Therefore, this will help government policymakers restore and maintain the ecological health of protected area communities at the landscape level.

5 Conclusion

Today, anthropological activities degrade forest lands; change climatic conditions and human intervention, and destroy wildlife and wildlife habitats. Many organizations and voluntary organizations work simultaneously to protect it nationally and globally. The SNP is a biodiversity in-situ genetic reservoir specially reserved for Indian tigers. We carried out an exploration of the relationship between forest health and climatic variability between 1972 and 2018 in protected areas using optical remote sensing. In SNP, the area was 1293 km2 in 1972 (dense forest, and open forest). The area was 1213 km2 in 1996. Then, 1287 km2 were covered in 2018. Therefore, there has been no significant change in the overall forest (dense and open) in the last 46 years. From 1972 to 1996, agricultural land increased by the rate of 1.02%; similarly, waterbodies increased by 2.61%; and barren land and settlement classes increased by 0.67%. From 1996 to 2018, agriculture increased by 1.01%; water bodies declined by 7.36%; and barren and settlement by 1.6%. Also, we used vegetative parameters such as NDVI, MSAVI-2, DVI, and MSI for forest health estimation. In the last 46 years, forest health has gradually increased as the NDVI significantly increased from these vegetative indices between 1972 and 2018. In 1972, there were higher MSAVI-2 values, which indicated a low contrast area of the soil after 1996, but in 2018, it was reduced to show a high contrast area of the soil. The DVI range in 1972 and 1996 has low values, indicates sparse regional vegetation, and is slightly enhanced in 2018. The MSI value in 1996 was higher compared to 2018. Temperature analysis of climatic factors has increased by one-degree Celsius in the last 46 years, and the annual rainfall rate has improved slightly from 1972 to 2010, after which it is constantly reduced. The SNP's forest health is positive from vegetative indices, and in 1972, we observed lower values of NDVI than in 1996 and 2018, and MSAVI-2 in 1972 and 1996 was more significant than in 2018. Therefore, this study provides a comprehensive overview of SNP forest health and plays a critical role in tiger conservation and management planning.