Spatiotemporal Analysis Land Use Land cover changes in South Kashmir Region of North-western Himalayas Using Landsat data

This paper presents a comprehensive analysis of land use changes in South Kashmir from 2000 to 2022, revealing signi�cant transformations in various land cover classes. Leveraging remote sensing and geographic information systems (GIS), the study examines the spatial patterns and temporal dynamics of land use and land cover, offering valuable insights into the region's landscape dynamics. Using supervised classi�cation techniques, satellite imagery was analyzed to identify 10 major land use classes. The �ndings demonstrate notable increases in horticulture and built-up areas, accompanied by declines in agricultural land, glaciers and snow, exposed rock, and water bodies. The expansion of horticultural lands, covering approximately 7% of the study area, has been attributed to the conversion of agricultural lands. This shift, coupled with the encroachment of settlements to accommodate the growing human population, has resulted in a substantial loss of approximately 757 km 2 of farmland agriculture, representing a total percentage change of about 13% during the study period. The implications of these changes extend beyond the local region, highlighting the urgent need for comprehensive and sustainable solutions to address human-induced challenges at a global scale. Furthermore, the study underscores the cost-effectiveness and e�cacy of geospatial technologies in conducting spatiotemporal analyses and formulating evidence-based policies for the sustainable management of natural resources. The insights gained from this study offer a solid foundation for informed decision-making and the development of targeted land management strategies in South Kashmir and other similar regions facing similar challenges.


Introduction
Land Use Land Cover is signi cant in all facets of human life since land is the fundamental source that satis es the needs of living things (Mandal et al., 2023).These studies aid in the establishment of baselines for LULC change studies, which are of great importance for the management and monitoring of the land surface (Congalton et al., 2014;Joshi et al., 2016).To properly monitor and analyse these changes, the development of sensor-based LULC datasets utilising a variety of satellite imaging techniques can be the correct approach (Pandey et al., 2019).Remote sensing methods have been identi ed as a strong tool to gain insight about the surface features of the Earth at various geographical and temporal scales (Liang et al., 2015;Satyanarayana et al., 2001).Changes in land use and land cover (LULC) are in uenced by a number of factors, primarily the expansion of agriculture and urban areas, deforestation, etc., and less frequently by factors like demographic and socioeconomic growth (Mandal et al., 2023.)Studies of land use and land cover (LULC) are crucial for comprehending land surface dynamics, hydrology, and how human activities interact with the environment.(Altaf et al., 2013;Meraj et al., 2013;Altaf et al., 2014).It is evident that land use and land cover are always changing across the multiple spatial scales.LULC variations are believed to be in uenced by both natural and human factors (Sarma et al. 2008).The rapidity and pattern of LULC change by humans is determined in terms of their socio-economic and political characteristics (Ojima et al. 1994).Recent LULC alterations triggered by human activity are discovered to be a signi cant factor in environmental deterioration, disrupting energy balance, chemical uxes, climate change, the greenhouse effect, eutrophication, deserti cation, ooding, acidi cation, and biodiversity loss (Jamal et al 2022); (Bae et al., 2015); (Awoniran et al. 2014; Sha q et al. 2019).As a result of the rapid and unchecked expansion of population, as well as economic and industrial development, it is unavoidable to assess LULC change in order to achieve a variety of social welfare objectives, especially in developing nations with rapid LULC changes (Dutta et al., 2019;Kumari et al 2019).In addition to promoting disaster risk reduction (DRR), LULC has a signi cant impact on climate change adaptation, due to the growing pressure on land resources caused by an aging population and urbanization (David et al. 2016;Shaw and Banba 2017).To address the increasing needs of the growing human population for basic sssrequirements and well-being, management methods must be chosen, planned, and put into practise using knowledge of land use and land cover.Further more information on LULC is essential for policymaking and improved decision making, managing land use plans sustainably, and gaining more insight into landscape changes (Pelorosso et al., 2009;Arveti et al., 2016;Lu et al.,2004;Seif and Mokarram, 2012).Researchers, stakeholders, and decision-makers require data from LULC to keep track of earth's resources.Additionally, this data enables them to assess the growth trends in various regions (Adeel 2010).Thus, it is crucial to have a comprehensive understanding of LULC and their associated changes is essential for mitigating and avoiding the various problems they may generate (Anderson, 2001).regions (Adeel 2010).Thus, it is crucial to have a comprehensive understanding of LULC and their associated changes is essential for mitigating and avoiding the various problems they may generate (Anderson, 2001).
During the past few years, a lot of progress has been made in LULC classi cation.It has been used for applications such as removing noise, masking cloud shadows, dividing images into regions, and identifying different land cover types (Afrin et al., 2019;Zhang et al.,2021).To monitor the changes and distribution of land use and land cover (LULC) from space, various methods have been devised.These include conventional ground-based mapping and satellite-derived mapping, which use the satellite data to identify and classify LULC types and trends (Talukdar et al., 2022).Satellite or aerial images offer diverse and valuable information for different regions, which can help to analyse the changes in land use over time (Wagner et al., 2013).By using remotely sensed data, a variety of approaches have been created and are being used for change detection to monitor changes in LULC, such as pixel-to-pixel comparison and post-classi cation comparison (PCC) (Lu et al. 2004).Remote sensing and Geographic Information Systems has emerged as the most accurate technologies in various elds for planning, mapping, assessing and evaluating spatial data on natural resources.(Tiwari et al., 2018;Bauer et al., 2005 ;Jamal, 2022).Using computational models and satellite data from remote sensing, researchers have been able to analyze LULC and nd out how it has changed over time and how it might change in the future (Li et al.,2014;Mourya et al.,2022)The development of new technologies in the elds of space exploration and optics has improved the accuracy of RS data enabling the collection of useful spatiotemporal data on LULC (Jensen, 2015).
Mountainous regions, such as Kashmir's Himalayas, frequently see LULC changes that are overly complex for the environment.Land use changes like this, which have detrimental long-term effects, are caused by intensi ed anthropogenic impact on the resources in these ecologically vulnerable locations (Rasool et al., 2016).Numerous scholars have undertaken extensive studies employing remote sensing techniques to examine the land use and land cover changes in Kashmir.Over the past three decades, the picturesque Kashmir valley has undergone signi cant transformations primarily attributed to population growth, economic expansion, alterations in agricultural practices, and the implementation of various development projects, (Alam et al. 2022;Amin et al. 2012).These studies where focused on land use dynamics in Srinagar city and Kashmir valley, they discovered that from 1990 to 2007, Srinagar saw substantial transformation.Additionally, the investigation demonstrated that changes in land use practises have led to the reduction of forests, open spaces, etc. (Lone et al 2018) draw the similar conclusion that the transition from dense to sparse forest was caused by the intense population pressure and found that built-up regions were taking over more and more plain areas at an alarming rate, displacing agricultural land.Sha q et al. ( 2021) while doing study on LULC changes in Kashmir himalayas pointed out that farmland agriculture has seen signi cant losses as a result of its conversion to horticulture and built-up areas.As a result, in this area without a national land use policy, it is necessary to ascertain the pace and trend of land cover/use conversion.While previous studies have provided valuable insights into LULC dynamics in the broader Kashmir valley, our research speci cally focuses on the south Kashmir Himalayan region.This area, although of immense ecological signi cance, has received relatively less attention in terms of LULC analysis.By unravelling the complex interplay of these driving forces, our research will contribute to a more comprehensive understanding of the socioeconomic and environmental implications of LULC changes in the south Kashmir Himalayas.

Study Area
South Kashmir Himalayas lies to the extreme north of India and stretches between the latitudes of 33°21′54′′N and 34°27′52′′N and the longitudes of 74°24′08′′E and 75°35′36′′E and covers an area of about 5454 km2.The study area has an elevation ranging from 1576 above mean sea level (amsl) to 5375 (amsl) (Fig. 1.) and the area is primarily exposed to Karewa deposits, also known as wudars.The region experiences a temperate climate with four distinct seasons: spring (March to May), summer (June to August), autumn (September to November) and winter (December to February).There are several perennial streams that drain the study area but Rambiara, Romushi, Lidder and Dudhganga are the most signi cant ones.These streams `signi cantly contribute to discharge of The Jhelum River, which has its origin close to Verinag.The south Kashmir consists of four districts viz Anantnag, Pulwama, shopian and Kulgam.According to the 2011 census, it has a total population of 2.38 million people.Anantnag town with an area of about 38.03 km2 is largest district of south Kashmir.However, over the past few decades, the study area has experienced unprecedented LULC changes, particularly in the extension of horticulture areas and urban settlements.

Data Acquisition and Preparation
This study employs remote sensing technology to comprehensively analyze land use and land cover (LULC) changes, leveraging the proven effectiveness of space-borne imaging.Landsat data from two distinct time periods, 2000 and 2022, were acquired and meticulously prepared for detailed analysis.
Landsat 5 (TM) data were utilized for 2000, while Landsat 8 (OLI/TIRS) data were selected for 2022.These datasets were sourced from the United States Geological Survey (USGS) Earth Explorer platform, providing a spatial resolution of 30 meters for precise evaluation of LULC dynamics in the study area.Stringent pre-processing techniques, were applied to enhance data quality and ensure accuracy.The acquired Landsat data were analyzed using software such as ERDAS and ArcGIS environment.A supervised classi cation approach using the Maximum Likelihood algorithm was employed in the ArcGIS platform to generate a land use class map for the study region.To ensure the accuracy of the classi cation results, the scheme was validated through ground truthing, by comparing the classi ed outputs with on-the-ground observations and reference data.Barren land, although a relatively smaller land cover class, also experienced changes during the study period.In 2000, barren land covered an area of 185.29 square kilometers, accounting for 3.40% of the total land area.By 2022, the area of barren land had slightly increased to 204.07 square kilometers, representing 3.75% of the total land area.This change of 18.78 square kilometers (0.35%) could be attributed to natural processes such as soil erosion, land degradation, or changes in land management practices.The persistence and slight expansion of barren land emphasize the need for sustainable land management strategies to mitigate further expansion and promote land productivity.
Built-up areas experienced signi cant growth over the study period, indicating rapid urbanization and infrastructure development in the region.In 2000, built-up areas covered 54.89 square kilometers, accounting for 1.01% of the total land area.However, by 2022, the area of built-up land had expanded to 127.17 square kilometers, representing 2.34% of the total land area.This remarkable increase of 72.28 square kilometers (1.33%) can be attributed to factors such as population growth, increased economic activities, and the establishment of residential, commercial, and industrial areas.The expansion of builtup areas underscores the need for effective urban planning, land-use regulations, and sustainable development practices to ensure e cient land utilization and mitigate the environmental impacts of urban expansion.
Exposed rock, another notable land cover class, underwent changes during the study period.In 2000, exposed rock covered an area of 947.85 square kilometers, accounting for 17.42% of the total land area.
By 2022, the area of exposed rock had decreased to 689.27 square kilometers, representing 12.66% of the total land area.This change of -258.58 square kilometers (-4.75%) could be attributed to natural processes such as weathering and erosion or human activities such as mining and quarrying.The decline in exposed rock area highlights the dynamic nature of landforms and geologic processes in the region.
Forests witnessed signi cant growth between 2000 and 2022, indicating efforts towards conservation and reforestation.In 2000, forests covered an area of 1,026.61square kilometers, accounting for 18.86% of the total land area.However, by 2022, the area of forested land had expanded to 1,301.37 square kilometers, representing 23.91% of the total land area.This increase of 274.76 square kilometers (5.05%) can be attributed to afforestation initiatives, conservation measures, and increased awareness about the importance of forests in ecosystem services, biodiversity conservation, and climate change mitigation.The expansion of forested areas highlights positive land-use practices in the region.
Glaciers and snow, an important land cover class in the region, underwent signi cant changes between 2000 and 2022.In 2000, glaciers and snow covered an area of 389.00 square kilometers, accounting for 7.15% of the total land area.However, by 2022, the area of glaciers and snow had decreased to 252.17 square kilometers, representing 4.63% of the total land area.This decline of -136.83 square kilometers (-2.51%) could be attributed to various factors such as climate change, glacial retreat, and altered precipitation patterns.The decrease in glaciers and snow highlights the vulnerability of these ecosystems to environmental changes and underscores the need for effective measures to mitigate their loss.
Grasslands, an important land cover class supporting diverse ecosystems, also experienced changes during the study period.In 2000, grasslands covered an area of 469.15 square kilometers, accounting for 8.62% of the total land area.By 2022, the area of grasslands had increased to 537.21 square kilometers, representing 9.87% of the total land area.This growth of 68.06 square kilometers (1.25%) indicates potential shifts in land management practices, natural succession, or changes in grazing patterns.The expansion of grasslands highlights the importance of preserving these ecosystems and promoting sustainable land-use practices.
Horticulture, an important land cover class associated with fruit and vegetable cultivation, witnessed signi cant changes over the study period.In 2000, horticultural areas covered 817.42 square kilometers, accounting for 15.02% of the total land area.However, by 2022, the area of horticulture had expanded to 1,236.59 square kilometers, representing 22.72% of the total land area.This remarkable increase of 419.17 square kilometers (7.70%) can be attributed to factors such as agricultural diversi cation, increased horticultural production, and changing market demands.The expansion of horticultural areas indicates the region's efforts to enhance agricultural productivity and capitalize on market opportunities.

Conclusion
The land use change analysis for South Kashmir between 2000 and 2022 revealed signi cant transformations in various land cover classes.The region witnessed declines in agricultural land, glaciers and snow, exposed rock, and water bodies, while experiencing expansions in built-up areas, forests, grasslands The assessment of LULC changes provides valuable information about land use and land cover patterns and processes.Both the remote sensing and GIS delivers a exible tool for assessing and analysing the spatial data necessary for change detection.In the present study supervised classi cation technique for 10 major classes was performed.The ndings of the study revealed that two main classes (Horticulture and Built up) have increased rapidly in the study area.The positive changes in these land use classes have caused negative changes in other land classes.The horticultural lands have been increased almost 7% during the period 2000-2022.The increase of horticultural lands can be attributed to its tremendous transformation into agricultural lands.With a loss of around 757 km2 and a total percentage change of about 13% from 2000 to 2022, farmland agriculture has seen severe losses.A sizable portion of agricultural lands are also being transformed to settlement areas to meet growing need of human populations.This study is crucial to nding a comprehensive, sustainable solution to these human-caused challenges that affect not just a speci c region but the entire world.Furthermore, the study illustrates that geospatial technologies provides a cost-effective tool for spatio temporal analysis and in the formulation of policies for sustainable management of natural resource Declarations Funding: No funding was received to conduct this study.
Ethical Approval: All the ethical standards of research publishing were taken care of during this study.
Con ict of Interest: The authors hereby declare no known con ict of interest in the research work reported in the manuscript.Showing NDVI map of South Kashmir for year 2000 and 2022.

Figure 1 Showing
Figure 1

Figure 2 Showing
Figure 2

Figure 3 showing
Figure 3

Figure 4 showing
Figure 4

Figure 5 Showing
Figure 5

Table 1
(Shahfahad et al., 2022;Talukdar et al., 2022)uce reliable LULC maps.The classi cation process identi ed nine distinct LULC classes, encompassing built-up areas, cropland, forest land, grassland, shrub, barren land, exposed rock, horticulture, snow-covered regions, and waterbodies.To address any initial classi cation errors, we employed post-classi cation techniques, including ground truths derived from classi ed scenes.These measures were undertaken to re ne and enhance the accuracy of the LULC classi cation outputs, thereby contributing to a more robust and precise analysis of land use and land cover dynamics within the study area.cation of land cover classes(Anderson, Hardy, Roach, & Witmer, 1976).Several techniques, including the Kappa coe cient, and error matrix, have been utilized in previous studies to assess the accuracy of LULC maps(Shahfahad et al., 2022;Talukdar et al., 2022).In this study, we employed the Kappa coe cient technique to evaluate the accuracy of the LULC maps of South Kashmir, using a sample of 200 randomly selected points.Ground observations for 2000 were acquired from Google Earth Pro due to the unavailability of eld data for those years.For 2022, a combination of eld visits and Google Earth Pro was utilized, particularly in areas with limited access or no eld data.The accuracy assessment results revealed an overall accuracy level of 83.39 percent, 88.79 percent for the years 2000, and 2022, respectively3.Results and DiscussionsLand Use Land Cover (LULC) change represents a signi cant and visually prominent transformation of the Earth's surface, drawing attention from researchers globally.The increasing trend of LULC change in watersheds re ects the global signi cance of economic factors driving human-induced modi cations to land.Previous studies have extensively assessed LULC changes in various catchments within the By examining the area and percentage of each land cover class for both time periods, as well as the change in area and percentage, a comprehensive understanding of the region's land use dynamics can be gained.The subsequent paragraphs present a comprehensive breakdown of the analysis conducted for each individual land use class: Scrub land, characterized by low vegetation and shrubs, also underwent notable changes during the study period.In 2000, scrub land covered an area of 138.43 square kilometers, accounting for 2.54% of the total land area.By 2022, the area of scrub land had increased to 446.25 square kilometers, representing 8.20% of the total land area.This signi cant growth of 307.82 square kilometers (5.66%) could be attributed to factors such as land abandonment, changing land management practices, or natural successional processes.The expansion of scrub land highlights the ecological dynamics and potential land-use transitions in the region.Agriculture, a signi cant land cover class in South Kashmir, experienced substantial changes between 2000 and 2022.In 2000, agricultural land covered an area of 1,393.88square kilometers, accounting for 25.61% of the total land area.However, by 2022, the area under agriculture had decreased to 636.67 square kilometers, representing only 11.70% of the total land area.This decline of 757.21 square kilometers (-13.91%)can be attributed to factors such as urbanization, changing agricultural practices, and land conversion for other purposes.The decrease in agricultural land highlights the shifting dynamics in the region's agricultural sector.
(Manandhar et al, 2009)LC) classi cation is a widely adopted approach for extracting valuable information from satellite imagery.In this study, we utilized False Color Composite (FCC) imagery to generate accurate LULC data speci cally for the South Kashmir.Among the array of available LULC classi cation techniques, we opted for a supervised classi cation method employing the Accurate assessment of the land use/land cover (LULC) classi cation plays a crucial role in evaluating the reliability and suitability of the generated maps for speci c purposes(Manandhar et al, 2009).It is important to quantify the classi cation accuracy to ensure the quality and interpretability of the maps produced.Generally, a minimum accuracy threshold of 80 percent is considered necessary for effective interpretation and identi Kashmir valley, revealing notable changes across these regions (Alam et al., 2019; Saha et al., 2022).It is commonly observed that transitions in LULC occur within Kashmir's catchments, transitioning from unirrigated systems such as forests and barren land to irrigated systems driven by agricultural and horticultural activities, which promote enhanced crop production and economic bene ts (Ritse et al., 2020; Singh and Panday, 2021).Over the past 22 years, substantial LULC changes have also been observed in South Kashmir, affecting various category classes across the entire region.These changes have primarily been concentrated within the selected study areas, attributed to the establishment of new settlements.The land use change analysis for South Kashmir between the years 2000 and 2022 reveals signi cant transformations in various land cover classes.The study area comprises 10 diverse land cover types, including (1) Water, (2) Scrub Land, (3) Horticulture, (4) Grasslands, (5) Snow and Glaciers, (6) Exposed Rock mass, (7) Built-up, (8) Barren Land, (9) Agriculture and (10) Forests for years 2000 and 2022 as depicted in Fig. 2 and Fig. 3. in hydrological patterns, water management practices, or human interventions.The decrease in water bodies emphasizes the importance of water resource management, conservation efforts, and maintaining ecological balance in the region.

Table 2
land use change statistics and change detection.The Figure.presents the inter-annual variation in NDVI values for South Kashmir using the MODIS dataset from 2000 to 2022.The analysis reveals a range of NDVI values, with the maximum ranging from 7,264.75 in 2000 to 7,667.33 in 2023, indicating the highest vegetation density observed in each respective year.Conversely, the minimum NDVI values range from 2,461.03 in 2000 to 3,160.74 in 2023, representing periods of reduced vegetation density.The range of NDVI values uctuates annually, re ecting changes in vegetation productivity and health.The standard deviation values range from 1,004.65 in 2007 to 1,166.17 in 2003, indicating the variability or spread of the data around the mean.These ndings demonstrate the dynamic nature of vegetation patterns in South Kashmir and the potential in uence of various factors such as climate, land use changes, and human activities.Understanding these variations is essential for effective land management, conservation strategies, and sustainable development in the region.