1 Introduction

According to Bender et al. (2005) and Mendoza et al. (2011), land use and land cover are two factors that are crucial in identifying the terrestrial environment in relation to both natural and human activities. Changes in land use and land cover have become an important component of modern techniques for monitoring environmental changes and managing natural resources (Tiwari and Saxena 2011; Rawat et al. 2013; Kumar et al. 2014). The land use of a particular region is determined by the interaction of land potential with cultural backgrounds, human requirements, and the state of society (Young et al. 1990; Turner et al. 1995; Bie et al. 1996; Karwariya and Goyal 2011; Tsai et al. 2019). Land cover identification provides a baseline from which monitoring activities (change detection) can be conducted, as well as providing ground cover data for baseline thematic maps (Rawat et al. 2013). Land use/cover change is a large and rapidly expanding process that is primarily driven by natural events and anthropogenic activities, resulting in changes that have an impact on natural ecosystems (Ruiz-Luna and Berlanga-Robles 2003; Turner and Ruscher 2004). The impacts of land use land cover (LULC) on ecosystem sustainability have become more important concerns in global change analysis. Human actions appear to be the most responsible for the current state of the Earth’s surface. Changes in the ground cover affect the balance of energy, water, and geochemical fluxes at the local, regional, and global levels, and these changes will unavoidably affect the sustainability of natural resources and socioeconomic activity (Vescovi et al. 2002). With increased population expansion, pressure is placed on a country’s limited natural resources, leading to changes in land cover. The processes of land cover and land use do not always imply changes brought about by socioeconomic factors or degradation of the land. In addition to having negative impacts on the biological community and the environment, socioeconomic reasons like increased urban activity and industrialization change the features of the landscape (Veeraswamy et al. 2017; Raju et al. 2018). Nowadays, land cover changes are regularly monitored using remote sensing (RS) in addition to satellite-based systems that collect geographic information periodically using a geographic information system (GIS) procedure. It enables research on the data that produces diverse spatial modelling, therefore optimizing the planning process as a whole.

Current tactics for managing natural resources and tracking environmental changes focus on increasing changes in land use and land cover. Studying Earth from space is now essential to understanding how man has affected Earth’s natural resource base over time. Space observations of Earth offer unbiased data on how people have used the land, especially in cases when land use is changing quickly and frequently without documentation. In order to manage natural resources, map the Earth’s features and infrastructure, and track environmental change, data from Earth sensing satellites has become increasingly important in recent years (Zubair 2006).

There are several procedures for identifying and assessing LULC. Among them, remote sensing and GIS techniques were commonly used by researchers in the field of LULC research (Nemani and Running 1997; Zhan et al. 2002; Mallick et al. 2008; Dewan and Yamaguchi 2009a, b; Wang et al. 2009; Mamun et al. 2013). Remote sensing, in combination with a Geographic Information System, is essential for acquiring sufficient data on LULC changes (Tiwari et al. 2018). LULC changes have previously been assessed using traditional surveys and assessments. However, satellite-based remote sensing and GIS are regarded as more useful and advantageous in terms of cost and time savings. It also gives large-scale information on LULC alterations as well as information on their geographic range (Hathout 2002; Yuan et al. 2005; Serra et al. 2008).

During the last three decades, the Kashmir Valley has seen substantial changes in land use and land cover as a result of increased population numbers, economic expansion, changes in agricultural techniques, and the implementation of various development projects. However, LULC changes are not being tracked in a systematic manner, and estimates of the scale of the changes are rare; as a result, information on the LULC dynamics in the region is limited. Using multi-temporal remote sensing satellite data, we attempted to map and quantify the LULC changes in the HWS Western Himalayas over the last three decades (1992–2006–2021). The findings of this study are expected to be highly informative for planners, resource managers, and policymakers for resource management.

2 Material and methods

2.1 Study area

A 341 km2 Hirpora wildlife sanctuary (figure 1) is located in the Pir Panjal range, 70 km southwest of Srinagar, the summer capital. It lies in the Shopian district of Kashmir at 33°39′55″N latitude and 74°39′40″E longitude at an altitude of 2546 m above mean sea level. The sanctuary’s geographical features contribute significantly to its unique ecosystem, making it a haven for a wide variety of plant and animal species (Ahmad et al. 2011). The altitude, climate, and topography create a habitat that supports the coexistence of numerous flora and fauna. The sanctuary is home to a rich array of plant life, with a diverse range of vegetation that includes dense forests, meadows, and alpine pastures. Tall coniferous trees such as pine and cedar dominate the landscape, providing shelter and sustenance to a variety of wildlife. The sanctuary’s unique flora also includes medicinal plants, adding an extra layer of ecological significance to the region. Hirpora Wildlife Sanctuary boasts an impressive diversity of wildlife. The sanctuary provides a refuge for several endangered species, including the Hangul deer and the Himalayan black bear. Avian enthusiasts will find delight in the numerous species of birds that call the sanctuary home, with migratory birds adding to the vibrant tapestry of the region during specific seasons. The presence of predators like leopards and various species of wild cats showcases the robustness of the sanctuary’s ecosystem (Ahmad et al. 2011).

Figure 1
figure 1

Location map of the study area.

Full size image

2.2 Methods

Remote sensing satellite data are essential for assessing land use land cover changes over time in an area and for mapping land-use patterns. For assessing LULC changes that have been in the study area, satellite images of Landsat-5 Thematic Mapper (TM) of 1992, Landsat-7 enhanced Thematic Mapper (TM) 2006, and Landsat-8 Operational Land Imager (OLI) 2021 (table 1) with a 30m resolution satellite imagery (Rasool et al. 2021) were used. Arc GIS software was used to prepare the study area boundary. After preprocessing and enhancement procedures, a supervised maximum likelihood classification technique was used to classify these images from 1992, 2006, and 2021. To find these categories of classification, a mix of fieldwork, map analysis, and firsthand experience with LULC sorts was used (Jensen 1986). Image processing was done by an earth resources data analysis system (ERDAS) Imagine 14 software. Eleven LULC classes were chosen based on a classification scheme developed by NRSC (National Remote Sensing Centre), Hyderabad, that was later accepted by the Department of Remote Sensing and Environment J and K in 2006. These classes are: (1) dense forests, (2) moderately dense forests, (3) scrub, (4) built-up, (5) barren/rocky, (6) agriculture/horticulture, (7) river bed, (8) snow cover, (9) alpine lakes, (10) glacier, and (11) grassland (table 2). The classes were improved with various colour composites such as Infrared, near Infrared, and red to identify different classes. An overlay procedure in ArcGIS was used to compute the LULC change transition matrix to determine the area that changed from one LULC category to another during the study period. The mixed pixel problems were later resolved using visual interpretation. Post-classification field trips were carried out to further refine the LULC classification and ground validate ambiguous places using GPS and other tools. In order to estimate the LULC change, multitemporal raster layers (1992, 2006, and 2021) were created and the accompanying data were compared.

Table 1 Characteristics of the primary/satellite data used in the present study.
Full size table
Table 2 Land use land cover classification scheme.
Full size table

The area coverage from the first and second years was subtracted to get the magnitude change for each land use class, as shown in equation (1).

$${\text{Magnitude }}={\text{ Magnitude of new year }}{-}{\text{ Magnitude of previous year}}.$$
(1)

After that, the percentage change (trend) for each type of land use was determined by multiplying the result by 100 and dividing the magnitude change by the base year (the first year), as shown in equation (2).

$${\text{Percentage change }} = \frac{{{\text{Magnitude of change }} \times 100}}{{\text{Base year}}}.$$
(2)

The difference between the final year and the initial year, which indicates the magnitude of change between corresponding years, was divided by the number of study years, i.e., 1992–2006 (15 yr), 2006–2021 (15 yr), and 1992–2021 (30 yr), respectively, to obtain the annual rate of change for each land use type using equation (3).

$${\text{Annual rate of change }} = \frac{{{\text{Final year}} - {\text{Initial year}}}}{{{\text{No}}.{\text{ of years}}}}.$$
(3)

3 Results

The study focused on Landsat TM imagery for 1992 to determine the past land use pattern in the study area. A cross-examination was used to verify if a past identified land use pattern exists or has been converted into another. Later, ground truth knowledge was conducted to identify each specific pixel and colour tone.

A detailed description of land use and land cover changes within the study area over the selected time period (1992–2006–2021) are depicted in figure 2(A–C). Table 3, along with figure 3, illustrates the changes in LULC classes from 1992–2021, along with the proportionate area covered for each land class. During 1992, snow cover (116.39 km2) was the most dominant land use class category, followed by grassland (73.94 km2) and barren (57.89 km2) out of the total area (341 km2) of the sanctuary. All land types showed trend changes from the year 1992–2020. Based on the data reported in table 3 and figure 3, each LULC category is represented by the area of the LULC category for three different years (1992, 2006, and 2021) grouped into 10 different classes. Most of the land cover classes showed an increase in area.

Figure 2
figure 2

Land use/cover status of the Hirpora Wildlife Sanctuary in (A) 1992, (B) 2006, and (C) 2021.

Full size image
Table 3 Category-wise land use distribution of HWS from 1992–2021.
Full size table
Figure 3
figure 3

LULC change graph for 1992, 2006 and 2021 at HWS.

Full size image

Grassland was the largest of all the identified classes in the study area. In 1992, grassland covered about 21.92% of the total area of the sanctuary, which increased to 32.43% in 2006 and 43.38% in 2021 (table 3). The class showed a net positive change of 21.46% from 1992–2021. However, the biggest change (10.95%) was observed between 2006 and 2021 (table 4 and figure 4). The growth of this class is attributed to the conversion of barren areas, which is contributing to its positive change. Scrub also registered an increase in its area from 8.73% (1992) to 9.23% (2006) and later jumped to 14.14% in 2021 with an annual positive rate of 5.41% from 1992–2021 (table 4). Agriculture/horticulture showed a positive increase from 0.15 km2 in 2006 to 0.19 km2 (table 3) with a positive annual rate change of 0.19% from 1992–2021 (table 4). Compared to the area of other classes in 1992, there has been a slight increase in the area of land under the river bed from 1.35, 1.40, and 1.57% in 1992, 2006, and 2021, respectively (figure 3).

Table 4 Landuse change assessment of HWS based on time frame data (1992–2021).
Full size table
Figure 4
figure 4

Changes within each land use and land cover class (1992–2001–2015).

Full size image

Additionally, the build-up class has experienced an increase in land area from 1.02 km2 (2006) to 1.17 km2 in 2021 (table 3 and figure 3) with a positive annual rate change of 0.35% (table 4).

Barren constitutes one of the largest LULC classes in this classification. Over the past 29 years, this category has undergone continuous positive change. The area under barren cover increased from 17.16% in 1992 to 24.24% in 2006 and 28.59% in 2021. From 1992 to 2006, the annual positive rate stands at 7.08% and continues to grow by 4.35% from 2006–2021 (table 3), with a net annual positive rate change of 38.53% from 1992–2021 (table 4). Moderately dense forest depicts one of the smaller classes of this classification. In 2006, 1.17% of the area was covered by moderately dense forests, which increased to 3.33% in 2020 (table 3). From 1992–2006, it has increased at a positive rate of 1.02%. The LULC class has shown a further increase of 1.14% between 2006 and 2021, with a net positive annual rate change of 4.50% from 1992–2021 (table 4).

Four classes showed a negative growth rate from 1992–2021, i.e., dense forests, river beds, alpine lakes, glaciers, and snow cover. After observations, it was found that during the period 1992–2021, the area considered well suited for snow cover shares had decreased from 34.51, 19.97, and 0.75% in 1992, 2006, and 2021, respectively (table 3); however, the highest reduction rate (19.22%) was recorded between 2006 and 2021 (table 4) with a negative net rate change of 33.76%. Dense forest land declined continuously throughout the study period (table 3). The area under forests decreased from 13.19% in 1992 to 8.76% in 2006 and 6.39% in 2021, respectively (table 3). Forest land decreased by 4.43% from 1992–2006, by 2.37% from 2006–2021, and there was a net reduction of 6.80% from 1992–2021 (table 4). The decrease in forests can be attributed to the rapid growth of the human population and the ever-increasing demand for fuel wood, timber, firewood, and other forest products. With a negative net rate change of 0.43% (table 4) and a total change of 1.43 km2 (table 3), the area of land covered in glaciers decreased from 0.36% in 1992 to 0.07% in 2021. Alpine lakes also showed a decrease in the area from 1992–2021. The area under alpine lakes decreased from 2.48 km2 in 1992 to 2.20 km2 in 2021. The highest reduction rate 0.05% was recorded between 1992 and 2006 with a negative net rate change of 0.08%.

3.1 The ‘From-To’ change (1992–2021)

To study the exact changes in each land use class, we created a change transformation matrix by comparing the 1992 raster cell to the 2021 (figure 5). The matrix table shows the areas that have changed, the areas that have remained unchanged, and the total areas that have changed in each class. The row values represent the class values in the culminating year of study (2021), while the column values represent the data from the initial year (1992). Table 5 and figure 6 given below illustrate how different land use classes have been transformed into others and different gains and losses in each land use/land type (in km2). As a result of the various land use classes contributing to net gains in grassland areas between 1992 and 2021, 22.91 km of barren lands were converted to grassland, and 19.81 km of scrubland and 8.43 km of dense forest land were converted to grassland areas. During this period, 59.46 km of snow cover changed into barren, 39.76 km into grassland, and 14.19 km into scrub, respectively. It was evident that there were significant changes and transitions among the various types of LULC as a result of the LULC change analysis. The most remarkable change between 1992 and 2021 has been in grassland and snow cover areas. As shown in table 5 and figure 5, this study presents the changes in land use/land type (in km2) and net changes to LULC classes over periods 1, 2, and 3. Graphs 6 and 7 show the changes in the four LULC classes and their unchanged areas during the study period.

Figure 5
figure 5

Change detection map between LULC of 1992 and 2021 between different classes. Here A/H = Agriculture/Horticulture, BR = Barren/Rocky, BU = Built-up, DF = Dense Forest, MD = Moderately Dense Forest, SR = Scrub, SC = Snow cover, RB = River bed, AL= Alpine Lake, GR = Grassland and GL = Glacier.

Full size image
Table 5 LULC transformation matrix from 1992 to 2021 between different classes.
Full size table
Figure 6
figure 6

Losses and gains in LULC (in km2) and no change contributions during 1992–2021.

Full size image

4 Discussion

Based on the results of the study period (1992–2021), most of the forest area has been converted into grassland and scrub. There has been an increase in the amount of grassland and barren land in the past two decades, possibly at the expense of forest cover (Olsson et al. 2000; Basha et al. 2018; Chamling and Bera 2020). The shift in forest areas may signal a complex interplay of factors affecting forest cover. This could include deforestation, land-use changes, or natural disturbances (Waikhom et al. 2023). The decline in dense forests over the study period from 1992 to 2021 suggests a significant change in forest cover. This decline could be attributed to various factors, many of which are associated with human activities and environmental changes (Tariq et al. 2022). Natural factors, as well as population pressure and human demands for timber, fuel wood, land for settlements, roads (Mughal Road), agriculture, etc., have led to an overall decline in forest area (Sharma and Roy 2007; Farooq and Rashid 2010). These human-induced factors are often interconnected and can have cumulative impacts on forest ecosystems. The challenge lies in finding a balance between meeting human needs and ensuring the sustainable management of natural resources, including forests. Efforts to address these challenges typically involve the implementation of conservation strategies, sustainable forestry practices, and policies that promote responsible land use. Community engagement and awareness are also crucial components of successful conservation and sustainable development initiatives (Bhat et al. 2022). The positive relationship between increasing green space buffer zones and participation suggests that environmental and aesthetic considerations play a role in influencing people’s decisions to participate in conservation (Kordani et al. 2022). Additionally, reforestation and afforestation programs aim to restore and increase forested areas to counteract declines in forest cover (Rasool et al. 2021). Moderately dense forests have shown an increase of 3.33% with a net rate increase of 4.50%. The increase in moderately dense forests within the Hirpora Wildlife Sanctuary can be influenced by various factors, and these changes may be a result of natural processes, conservation efforts, or a combination of both. However, these results are contrary to Sen et al. (2002) findings but are in conformity with the study of Yadav et al. (2012). Agriculture/horticulture increased from 0.15–0.19 km2. This increase is attributed to the conversion of forests and woodlands to cropland (Houghton 1994; Williams 1994; Kaul et al. 2009). The increasing horticulture land area is due to people focusing on cash crops, and this has increased their socioeconomic status (Rigterink 1989; Uhlig and Kreutzmann 1995). The increase in agricultural, urban, and water body areas over the years reflects patterns of urbanization, agricultural expansion, and potential changes in hydrological features (Waikhom et al. 2023).

The slight increase in river bed area from 1.35% in 1992 to 1.57% in 2021 suggests a change in the landscape over the study period. Land transformation, influenced by various natural and anthropogenic factors like sedimentation and erosion, urbanization and infrastructure development, agricultural activities, climate change, and natural events, could indeed be a contributing factor to this increase in river bed area (Fazal and Amin 2011). The significant increase in barren or rocky land area from 57.89 km2 in 1992 to 96.42 km2 in 2021 suggests a notable transformation of the landscape. Several factors could contribute to this increase in barren or rocky land, such as land use changes, deforestation, natural processes, climatic factors, and land degradation. This study is in conformity with Kaul et al. (2009). The positive annual increase of 0.35% in the buildup class, observed from 1992 to 2021, indicates a gradual expansion of built-up areas over time. Such urban or built-up expansion can have significant implications for agricultural lands and wetlands (Kumar 2016; Hu et al. 2019). Despite the rise in settlements resulting from the increase in households, there has been an exponential increase in business establishments. The conversion of fertile and prime agricultural land to other uses is indeed a concerning trend with potential implications for food security, environmental sustainability, and rural economies (Salem et al. 2020). Several factors contribute to this phenomenon, including urbanization, infrastructure development, industrialization, and environmental impact. A study from Mahanadi Reservoir Complex, Chhattisgarh, India, sheds light on the intricate connections between land use changes and environmental dynamics, emphasizing the need for informed land-use planning, conservation strategies, and sustainable urban development practices to balance human activities with environmental considerations (Verma et al. 2023). The observation that LULC (land use and land cover) changes are increasing in the study area and that economic factors play a significant role aligns with the broader understanding of land-use dynamics globally. Economic considerations, driven by factors such as population growth, urbanization, and industrialization, often influence how land is utilized (Butt et al. 2015).

5 Conclusion

An overview of various land-use changes that have occurred in the study area is included in the present study. The findings indicate that forests and snow cover classes are continually decreasing in the study area, with a decrease of 21.3 and 39.9 km2 and a negative rate change of 6.25 and 11.7%, respectively. However, the majority of the classes showed an increase in the area from 1992 to 2020. The most visible of them was an increase in barren land (9.87%), followed by open scrub (3.42%) and moderately dense forests (2.22%). Agriculture, build, and willow plantation classes have shown a slight gain of 0.67, 0.7, and 0.14% in the area, respectively.

The statement ‘Human-prompted changes have irreversibly resulted in environmental degradation’ highlights a serious concern about the long-lasting and irreversible impact of human activities on the environment, as was demonstrated in the study. There is no doubt that the massive degradation of the Hirpora Wildlife Sanctuary, in particular, caused by a human is imminent. Sustainable management of Hirpora amidst decreasing forests involves practices such as reforestation, habitat restoration, community engagement, controlled tourism, and wildlife conservation initiatives. Implementing eco-friendly policies and promoting awareness about the importance of preserving biodiversity are crucial components. The suggested strategies for managing the land, establishing nurseries, promoting sustainable farming, and encouraging forest regeneration in the Hirpora Wildlife Sanctuary align with principles of sustainable development and conservation. Additionally, the call for enforcing the Jammu and Kashmir Wildlife (Protection) Act, 1978 underscores the importance of legal frameworks in preserving biodiversity and protecting wildlife. The success of these strategies depends on effective collaboration, continuous monitoring, and adaptive management. It also involves balancing conservation goals with the needs of local communities for sustainable and equitable development.