Harnessing LiDAR and SAR Remote Sensing for Climate-Adaptive Heritage Monitoring:A PRISMA-Based Review

Abstract

Climate change poses a significant threat to cultural heritage worldwide, manifesting through sea-level rise, increased humidity, extreme weather events, and shifting land dynamics. This review explores the potential of advanced remote sensing technologies, particularly Light Detection and Ranging (LiDAR) and Synthetic Aperture Radar (SAR), for climate-adaptive heritage monitoring. LiDAR provides high-resolution topographic and structural data, enabling accurate 3D reconstruction and vulnerability assessment, while SAR offers all-weather, day-and-night imaging for deformation monitoring and early detection of structural changes. Together, these technologies provide a robust, non-invasive framework for large-scale and continuous heritage observation. This Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-based review synthesizes recent advances in LiDAR and SAR applications for cultural heritage conservation, with emphasis on case studies from climate-vulnerable regions. It highlights interdisciplinary approaches integrating geospatial analytics, machine learning, and historical data to strengthen conservation strategies. The synthesis reveals a shift toward multi-sensor and multi-scale monitoring approaches, underscoring the complementary roles of LiDAR and SAR. Demonstrative case studies illustrate the empirical contributions: terrestrial laser scanning of the 18th-century Dutch House in Patna achieved sub-2 cm 3D model accuracy (Root Mean Squared Error [RMSE] = 1.3 cm) relative to independent total-station measurements, establishing a baseline for structural health monitoring, while a preliminary Persistent Scatterer Interferometry (PSInSAR) workflow over Shergarh Fort serves as a methodological proof-of-concept for SAR-based deformation screening. The novelty of this review lies in bridging methodological advances with practical implementation, particularly for data-constrained environments in the Global South, and in advancing a formalized, tiered integrated LiDAR-SAR framework for climate-adaptive heritage monitoring.

1. Introduction

Cultural heritage stands as a testament to human history, embodying the artistic, architectural, and social achievements of past civilisations (Silverman & Ruggles, 2007). However, this irreplaceable legacy is under threat from the pervasive impacts of climate change (Lu et al., 2025). Rising sea-levels, increased humidity, extreme weather events, and shifting land dynamics are accelerating the degradation of ancient structures, archaeological sites, and intangible cultural landscapes (Silverman, 2011). These changes are particularly acute in climate-vulnerable regions, where heritage sites often lack the resources and infrastructure for effective monitoring and +conservation (Hall et al., 2016).

In this context, advanced remote sensing techniques such as Light Detection and Ranging (LiDAR) and Synthetic Aperture Radar (SAR) have emerged as powerful tools for safeguarding cultural heritage (Cigna et al., 2024). LiDAR offers unparalleled precision in capturing high-resolution topographical and structural data, thus enabling 3D reconstructions, change detection, and vulnerability assessments (Li et al., 2023). SAR complements these capabilities with its all-weather, day-and-night imaging capability, making it invaluable for surface deformation analysis, flood mapping, and early detection of structural shifts (Tapete & Cigna, 2017). Together, these technologies provide a non-invasive, scalable, and continuous monitoring framework, essential for adapting heritage conservation strategies to the realities of a changing climate (Chen et al., 2018).

Despite their complementary strengths, LiDAR and SAR face distinct technical and operational limitations that underscore the necessity of integrated approaches (Li & Xiao, 2025). LiDAR, particularly terrestrial and airborne systems, delivers exceptional spatial resolution (millimeter to sub-meter) and accurate 3D structural models but is constrained by high equipment and operational costs, limited temporal frequency (typically single or infrequent surveys), susceptibility to atmospheric interference (e.g., fog, heavy rain), and significant data processing complexity, making it impractical for continuous monitoring across large or inaccessible sites (Pardini et al., 2019).

In contrast, SAR provides consistent, all-weather, day-and-night monitoring with high temporal resolution (e.g., Sentinel-1’s 6-day repeat cycle) and wide-area coverage, enabling systematic deformation tracking; however, its utility is hindered by lower spatial resolution (meter to tens of meters), sensitivity to temporal decorrelation over vegetated or rapidly changing surfaces, geometric distortions (foreshortening, layover, shadow) in complex terrain, and challenges in isolating atmospheric and noise contributions in interferometric analyses (Zhang et al., 2024). This divergence—LiDAR’s strength in high-detail structural documentation versus SAR’s advantage in continuous temporal monitoring—highlights that no single sensor can fully address the multi-scale demands of climate-adaptive heritage conservation, reinforcing the critical need for integrated, multi-sensor frameworks (Kazanskiy et al., 2025).

Despite their complementary strengths, LiDAR and SAR face distinct technical and operational limitations that underscore the necessity of integrated approaches (Li & Xiao, 2025). LiDAR excels in high-detail structural documentation but is constrained by cost, temporal sparsity, and atmospheric susceptibility (Pardini et al., 2019). SAR offers consistent, all-weather monitoring with wide-area coverage but has lower spatial resolution and can suffer from decorrelation and geometric distortions (Zhang et al., 2024). This divergence—LiDAR’s spatial precision versus SAR’s temporal continuity—motivates the development of unified, multi-sensor frameworks (Kazanskiy et al., 2025). However, existing research remains fragmented, often limited to single-sensor workflows, and a systematic synthesis of their climate-adaptive integration for heritage monitoring is lacking (Francini & Rozochkina, 2024). Furthermore, adoption is uneven, particularly in the Global South, where financial and technical constraints hinder deployment (Tapete & Cigna, 2019).

This review addresses these gaps by synthesizing advances through a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-based methodology and proposing an integrated LiDAR–SAR monitoring framework. At the same time, current studies are often fragmented, sensor-specific, and insufficiently aligned with climate-adaptive conservation needs. This review seeks to bridge this gap by synthesizing recent advances in the integration of LiDAR and SAR for cultural heritage conservation. Through a PRISMA-based methodology, this review aims to highlight interdisciplinary approaches that combine geospatial analytics, machine learning, and historical data to enhance conservation efforts. In addition, the review connects methodological progress with practical implementation challenges through representative application examples. By examining these studies from climate-vulnerable regions, this review underscores the urgent need for a paradigm shift in heritage conservation strategies and the application of advanced remote sensing technologies.

1.1. Impact of Climate Change on Cultural Heritage

Global climate change is one of the most significant threats to cultural heritage worldwide, exacerbating both gradual degradation and sudden destruction of ancient structures and archaeological sites (Chen et al., 2023). Extreme weather events resulting from Global climate change, such as flash floods, landslides, and wildfires, impact cultural heritage sites severely and often cause irreparable damage (Zhou et al., 2015). These impacts are not only increasing in frequency but also in spatial extent, posing systemic risks to heritage sites across diverse climatic regions. The cumulative effects of repeated exposure to such hazards further accelerate material deterioration and structural instability, necessitating continuous and multi-temporal monitoring approaches. The heritage sites located close to coastlines are highly vulnerable to sea-level rise and saltwater intrusion, which accelerates material decay and causes structural deterioration (Nazarnia et al., 2020). In such coastal environments, early detection of structural weakening and surface degradation is critical, highlighting the importance of integrating remote sensing technologies capable of capturing both surface and subsurface changes over time. The case of Dwarka in India, with a significant part of the ancient city submerged under the sea, is well documented (Gaur et al., 2004). Another example is the Thiruchendur Murugan temple in Tamil Nadu in India, which is located close to the coastline and is facing threats from climate change-induced sea-level rise in the near future (Trouillet, 2017).

The elevated humidity and temperatures promote biological growth such as mould and lichen and chemical reactions such as salt crystallisation leading to the deterioration of masonry, paintings, and organic materials (Stanaszek-Tomal, 2020). These processes often operate at micro-scales and may not be immediately visible, thereby requiring high-resolution monitoring techniques such as LiDAR to capture subtle structural and surface-level changes. The temperature stresses further negatively impact historic materials, engravings, and polishing, thus causing expansion, contraction, and cracking (Widomski et al., 2025). Intensified storms, heavy rainfall, and flooding can cause immediate physical damage such as collapse or erosion, while prolonged droughts can damage the foundations in arid environments (Othman et al., 2023). The dynamic and multi-hazard nature of these climate-induced impacts underscores the need for integrated monitoring frameworks that combine structural mapping with deformation analysis, which can be effectively achieved through the complementary use of LiDAR and SAR technologies.

1.2. Application of LiDAR Technology for Heritage Conservation

LiDAR has emerged as an important tool in heritage conservation that offers high-resolution 3-dimensional (3-D) mapping of cultural heritage sites with unparalleled accuracy (Kantaros et al., 2023). Its ability to capture fine-scale geometric details makes it particularly suitable for documenting structural conditions, detecting minor deformations, and supporting restoration planning. The working principle of LiDAR involves emitting Laser pulses towards a target and measuring the time-delay of reflected signals for distance calculations (Li et al., 2024). The Time of Flight (ToF) principle is governed by Eq. (1) (Killinger, 2014).

\(\begin{align}d=\frac{c . \Delta t}{2}\end{align}\)       (1)

where d is the distance to the target, c is the velocity of the laser pulse (3 × 108 m/s), and Δt is the time taken for the round trip of the Laser pulse. By scanning a surface with thousands to millions of laser points, LiDAR generates dense point clouds that represent the geometric and topographical features of the heritage structures (Yang et al., 2023). However, despite its high spatial accuracy, LiDAR is often limited by high operational costs, data processing complexity, and limited temporal frequency, restricting its use for continuous monitoring applications. These point clouds can be processed to create detailed 3-D models, enabling precise documentation, structural health monitoring, and vulnerability assessments (Wang et al., 2018). The Airborne LiDAR Scan (ALS) and Terrestrial LiDAR Scanning (TLS) are the two primary modalities used in heritage 3-D model development applications (Liu et al., 2025). ALS is particularly effective for large-scale surveys, such as uncovering hidden archaeological features beneath dense vegetation, while TLS provides millimeter-level accuracy for intricate architectural details (Strzałkowski et al., 2025).

LiDAR’s applications in cultural heritage conservation are multifaceted. It facilitates the structural deformation through comparative multi-temporal scan analysis, such as the Multi-scale Model TO Model Cloud Comparison (M3C2) method (Saiti et al., 2025). The M3C2 model computes the local surface distances between point clouds defined by Eq. (2) (Yang & Schwieger, 2023).

\(\begin{align}D_{M 3 C 2}=\frac{\left(p_{2}-p_{1}\right) . n_{1}}{\left\|n_{1}\right\|}\end{align}\)       (2)

where p₁ and p₂ are the corresponding points in two scans, and n₁ is the normal vector at p₁. This allows the conservators to identify the subtle cracks and shifts in heritage structures caused by climate-induced stressors like ground subsidence or thermal expansion (Cavalagli et al., 2019). Additionally, LiDAR aids in flood risk modelling by generating Digital Elevation Models (DEM) that simulate water flow and inundation scenarios, critical for safeguarding low-lying heritage sites (Trepekli et al., 2022). Machine Learning (ML) techniques such as Convolution Neural Networks (CNNs), further enhance LiDAR data processing by automating feature extraction (such as identification of corroded walls or collapsed facades) (Meira et al., 2024).

1.3. Application of SAR Remote Sensing for Heritage Conservation

SAR is a powerful remote sensing technology that provides all-weather, all-time monitoring capabilities, making it indispensable for heritage conservation in climate-vulnerable regions (Tripathi and Kumar, 2018). Unlike LiDAR, SAR enables continuous temporal monitoring and is particularly effective in detecting surface deformation; however, its relatively lower spatial resolution and sensitivity to decorrelation in complex environments can limit its effectiveness for detailed structural mapping. Unlike the optical sensors, SAR operates by emitting microwave pulses towards the earth and measuring the backscattered signals, which are influenced by surface roughness, moisture and structural properties (Tripathi et al., 2018). The radar equation that governs the SAR signal strength is given by (Mahto et al., 2025).

\(\begin{align}P_{r}=\frac{P_{t} G_{t} G_{r} \lambda^{2} \sigma^{0}}{(4 \pi)^{3} R^{4}}\end{align}\)       (3)

where P_r is the power received, P_t is the power transmitted, G_r and G_t are the receiving and transmitting antenna gains, λ is the wavelength, σ⁰ is the radar cross section (a measure of reflectivity), and R is the distance between the sensor and the target (Tripathi et al., 2018). While the underlying radar equations are fundamental to SAR signal interpretation, this review focuses primarily on application-oriented insights relevant to heritage monitoring rather than theoretical derivations. One of the key applications of SAR remote sensing is the surface deformation mapping and monitoring using SAR Interferometry (InSAR) (Tripathi and Kumar, 2018). The InSAR measures millimetre levels of displacements by analysing the phase differences between two SAR image acquisitions with the same sensor geometry over the same area, over a period of time (Reshi et al., 2023). The phase difference (Δφ) is related to ground displacement (Δd) as shown in Eq. (4) (Mahto et al., 2025).

\(\begin{align}\Delta \varphi=\frac{4 \pi}{\lambda} \Delta d\end{align}\)       (4)

This technique has been successfully applied to monitor subsidence in ancient fortifications, such as the Rohtasgarh fort in India, as shown by Tripathi and Kumar (2018), where gradual movements due to soil erosion and hydrological changes were observed. Another key area is flood mapping where SAR’s sensitivity to water surface (low backscatter values over waterbodies) helps in the assessment of water logging risks for low-lying heritage sites (Aziz et al., 2022).

1.4. Structured Comparison of Previous Studies

To strengthen the critical synthesis of existing literature, Table 1 presents a structured methodological synthesis of representative studies that employ LiDAR, SAR, or their integration for heritage monitoring. Studies were selected based on their methodological rigor, relevance to climate-adaptive conservation, and geographical diversity. The comparison (Table 4) evaluates each study across five dimensions: (1) primary technology, (2) climate-related focus, (3) methodological strengths, (4) key limitations, and (5) contribution to integrated monitoring frameworks.

Table 1. Methodological synthesis of integrated LiDAR–SAR studies

The structured comparison reveals three overarching trends. First, while LiDAR-based studies (Said & Shahrin, 2025; Pinheiro et al., 2024) excel in high-resolution structural documentation and defect detection, they are predominantly static or infrequent, limiting their utility for monitoring dynamic climate-induced changes. Conversely, SAR-based studies (Tang et al., 2016; Tripathi & Kumar, 2018) enable continuous deformation monitoring but suffer from lower spatial resolution and coherence challenges in vegetated or complex terrains. Second, only a few studies (Fu et al., 2026) have begun to operationalize multi-scale integration, yet most remain fragmented—either focused on a single sensor or lacking explicit linkage to climate variables. Third, geographical biases persist: the Global South is severely underrepresented, with most integrated studies concentrated in Europe and East Asia. This synthesis underscores the pressing need for unified frameworks that combine LiDAR’s spatial fidelity with SAR’s temporal regularity, incorporate climate data, and address accessibility disparities.

2. Decadal Review

The evolution of remote sensing for heritage conservation from 1990 to the present reflects a trajectory from isolated, single-sensor applications toward integrated, multi-scale monitoring frameworks. Understanding this progression is essential for contextualizing the current potential for LiDAR–SAR integration in climate-adaptive heritage monitoring. The following decadal analysis examines not only technological milestones but also the cumulative contributions of each period to the eventual synthesis of structural documentation (LiDAR) and temporal monitoring (SAR) capabilities.

2.1. 1990–2000: The Early Phase

The 1990s saw the pioneering integration of remote sensing into heritage conservation, primarily through aerial photogrammetry and early satellite data from the Landsat series (Giardino, 2011). Landsat’s multispectral data enabled the first multispectral archaeological prospection, revealing buried structures through crop marks in regions such as the Near East (Clark et al., 1998). Aerial thermography emerged as a novel technique, with notable applications at Stonehenge revealing sub-surface anomalies (Epperly et al., 1999). The development of early GIS platforms allowed for better spatial analysis of heritage sites, though manual digitization limited efficiency (Carrara et al., 1999).

The GPS technology became newly available for public use in this period enhanced survey precision for sites (Lewandowski et al., 1999). Experimental LiDAR prototypes such as NASA’s airborne topographic mapper demonstrated potential for digital elevation modelling, but costs restricted applications due to high budget (Eisele et al., 1999). This time-frame’s legacy lies in establishing remote sensing as a viable archaeological tool, through resolution and computational constraints hindered quantitative analysis (Blom et al., 2000). Overall, this period was characterized by technological experimentation and limited integration across sensing platforms, highlighting the early-stage nature of remote sensing applications in heritage conservation.

This foundational period established remote sensing as a viable archaeological tool, but more importantly, it revealed the inherent limitations of single-sensor approaches: optical imagery provided broad coverage but lacked structural detail; early LiDAR demonstrated topographic potential but was cost-prohibitive; and SAR remained largely untested in heritage contexts. These limitations set the stage for the sensor diversification that characterized the following decade.

2.2. 2001–2010: Sensing for 3-D Reconstruction

The 2000s revolutionised heritage documentation through airborne LiDAR, notably with the 2012 Cambodian Archaeological LiDAR Initiative (CALI) led by Damian Evans and colleagues, which uncovered extensive urban networks surrounding Angkor Wat, revealing a formally planned landscape extending far beyond the previously known temple complex (Evans & Moylan, 2013). Commercial TLS systems such as the Leica HDS3000 enabled millimetre accurate recording of monuments exemplified by the Digital Michelangelo Project (Bernardini et al., 2002).

The 2009 introduction of Structure from Motion (SfM) algorithms (Arana-Daniel et al., 2017), democratised photogrammetry with applications at Pompeii, generating textured 3-D models (Apollonio et al., 2021). Satellite SAR entered heritage monitoring, through ESA’s ERA-2, detecting the 2003 Bam earthquake impact on Persian ruins (Chini et al., 2009). Hyperspectral sensors such as Hyperion advanced material analysis, identifying pigment degradation in the Lascaux cave (Jal et al., 2025). Despite these advances, processing workflows remained manual; the 2006 Vienna memorandum (Eze et al., 2024) first advocated standardized digital documentation, highlighting persistent interoperability challenges (Lian & Xie, 2024).

The 2000s demonstrated that high-resolution structural documentation was achievable, but the temporal dimension of monitoring remained underdeveloped. While airborne LiDAR revolutionized landscape archaeology and TLS enabled millimetre-accurate recording of monuments, these surveys were typically single-acquisition ‘snapshots’ that could not capture ongoing change. Simultaneously, the first SAR applications in heritage (e.g., post-earthquake damage assessment) hinted at the technology’s potential for deformation monitoring, though integration with LiDAR-derived structural models had not yet been conceptualized. This decade thus established the complementary foundations—LiDAR for structure, SAR for change detection—that would later converge.

2.3. 2011–2020: Age of Machine Learning and Advanced Sensing for Heritage Conservation

Deep learning models transformed heritage conservation with CNNs automating feature detection with LiDAR data (Gallwey et al., 2019)-a breakthrough achieved in Maya settlement analysis (Gallwey et al., 2019). The 2014 launch of Sentinel-1 enabled free SAR data for InSAR monitoring and subsidence tracking at the Rome Colosseum (Grilli & Remondino, 2019). Unmanned Aerial Vehicle (UAV) photogrammetry became ubiquitous with projects such as the 2016 Palmyra arch reconstruction (Centre national d'études spatiales, CNES), achieving sub-centimetre accuracy (Kršák et al., 2016). Full waveform LiDAR, such as Riegl VQ-1560i penetrated the dense forest canopies in Amazonian archaeological study (Nikolakopoulos et al., 2017). Google Earth Engine (GEE) facilitated the global scale analysis, including a 2019 study quantifying looting damage in Mesopotamia (Daiyoub et al., 2023). The ARIADNE project (EU HORIZON 2020) established linked open data for heritage, though real-time data processing limitations persisted (Ioannides et al., 2017).

The 2011–2020 period marked a decisive shift toward operational integration. The availability of free, high-temporal-resolution SAR data from Sentinel-1 enabled systematic deformation monitoring at heritage sites, while advances in machine learning automated the processing of increasingly large LiDAR datasets. Critically, this decade saw the first explicit frameworks for multi-sensor heritage monitoring, with studies beginning to combine InSAR-derived displacement measurements with LiDAR-based structural models. The convergence of these capabilities—enabled by open data policies, computational advances, and growing climate urgency—directly sets the stage for the integrated LiDAR–SAR frameworks now emerging in the current decade.

2.4. 2021–2025: Notable Studies in the Recent Past

Building on the integration trends established in the previous decade, recent advances (2021–2025) have focused on operationalizing multi-sensor frameworks for climate-adaptive heritage monitoring. Key developments include digital twins for climate resilience, exemplified by the 2023 Notre-Dame reconstruction (Güner & Gül, 2025). Persistent Scatterer Interferometry (PSInSAR) time-series analysis of Venice’s basilicas (Rigamonti et al., 2023) achieved an estimation of 1mm/year deformation velocity. Multi-scale remote sensing approaches, integrating satellite imagery, UAV surveys, and geophysical methods, have enabled non-invasive subsurface imaging and characterization of Machu Picchu’s construction history, revealing previously undocumented phases of site development (Masini et al., 2023). The archiver project integrated multi-sensor data for predictive conservation, while ethical frameworks like the 2021 London charter 2.0 addressed Artificial Intelligence (AI) biases in data-based heritage conservation algorithms (Li et al., 2025). Persistent challenges include the Global South’s technology gap; a 2024 study by Eze et al. (2024) highlighted that fewer than 5% of African heritage sites utilize advanced monitoring technologies (Eze et al., 2024). Emerging solutions include low-cost radar microsatellites such as ICEYE and edge computing for real-time analysis (Li et al., 2024).

3. Review Methodology

This review employs the PRISMA framework to ensure methodological rigor, transparency, and reproducibility (Sarkis-Onofre et al., 2021). PRISMA is a widely recognised standard for systematic reviews, emphasizing structured literature selection, data extraction and synthesis (Page et al., 2021). The review methodology begins with a comprehensive search across multidisciplinary databases, including Scopus, Web of Science, IEEE Xplore, and Google Scholar, using keywords such as LiDAR, SAR, cultural heritage conservation, and climate change impacts (Kumar et al., 2025). The search was restricted to publications from January 1990 to December 2025 and to English-language records. The initial broad search was intentionally designed for high sensitivity to capture all potentially relevant studies, which accounts for the large initial yield (n=1,485). Grey literature (conference abstracts without full papers, theses, and non-peer-reviewed reports) was excluded during screening, as specified in the inclusion criteria.

The initial search yields a broad pool of studies, which are filtered through pre-defined inclusion and exclusion criteria (Table 2). Studies are included if they focus on LiDAR/SAR applications in heritage monitoring, explicitly address climate-related risks or have direct implications for climate-adaptive monitoring frameworks, and are published in peer-reviewed journals (including review articles) or peer-reviewed conference proceedings with full technical content between 1990 and 2025. Of the 70 included studies, 63 (90%) are peer-reviewed journal articles, 5 (7%) are peer-reviewed conference papers, and 2 (3%) are systematic reviews from indexed sources. Exclusion criteria omit non-English publications, non-empirical studies, and research lacking technical or contextual relevance.

Table 2. Inclusion and exclusion criteria applied during study selection

Overall, a total of 1,485 studies were initially identified through comprehensive database searches across Scopus (n=698), Web of Science (n=441), ScienceDirect (n=346), and supplementary searches in IEEE Xplore and Google Scholar. After removing 616 duplicate records using reference management software (Mendeley), 869 unique records remained for title and abstract screening. Following this initial screening, 795 records were excluded based on pre-defined eligibility criteria: 622 records did not focus on LiDAR or SAR applications in heritage monitoring, 55 were non-empirical studies (commentaries, editorials, or theoretical papers without case data), 29 were excluded due to language restrictions (non-English publications), and 89 records were excluded for lacking sufficient technical detail or relevance to climate-related heritage risks. The remaining 74 full-text articles were sought for retrieval and detailed eligibility assessment. Of these, 4 articles could not be retrieved due to access restrictions or incomplete bibliographic information. Consequently, 70 studies met all inclusion criteria and were included in the final qualitative synthesis.

The screening process involves title/abstract review followed by full text assessment with discrepancies resolved through at least three rounds of reading. Data extraction captures key metrics: study objectives, sensor types (example terrestrial LiDAR and Sentinel-1 SAR), analytical methods (example InSAR, M3C2), and case study outcomes. This review prioritises interdisciplinary studies integrating geospatial analytics, machine learning (example CNNs for LiDAR data), and historical climate data to highlight innovative approaches. To mitigate bias, methodological quality was assessed for included empirical studies (n=58) using a modified Newcastle-Ottawa Scale adapted for remote sensing case studies, evaluating three domains: (1) clarity of data acquisition and processing methods, (2) validation against ground-truth or independent measurements, and (3) transparent reporting of limitations. Studies were rated as high (score 7–9), moderate (4–6), or low (1–3) quality. Of the 58 empirical studies assessed, 42 (72%) were rated high quality, 14 (24%) moderate quality, and 2 (3%) low quality. Low-quality studies were retained to avoid selective reporting bias but are noted in the synthesis with appropriate caveats. The synthesis adopts a thematic analysis, categorising findings into technical advancements (example quantum LiDAR), regional applications (example flood mapping in India), and policy implications. Case studies such as SAR-based displacement analysis of Shergarh Fort and 3-D LiDAR mapping of Dutch House (Patna college) are analysed to demonstrate the practical utility.

The review also identifies gaps, such as the underrepresentation of Global South studies and proposed solutions like low-cost microsatellites to bridge technological disparities. The overall PRISMA review methodology is shown in Fig. 1, and the inclusion and exclusion criteria for literature in this review are summarised in Table 2. The relatively modest size of the final included corpus (n=70) relative to the initial identification pool (n=1,485) warrants explicit justification, as several factors contribute to this substantial reduction.

Fig. 1. PRISMA review methodology adapted in this study.

First, the review’s specific focus on the convergence of LiDAR and SAR technologies for climate-adaptive heritage monitoring represents a deliberately narrow scope. The initial search strategy employed broad keyword combinations (e.g., “LiDAR AND heritage,” “SAR AND cultural heritage”) to maximize sensitivity and avoid missing relevant studies, as recommended by systematic review guidelines (Lefebvre et al., 2022). This inclusive approach necessarily captures a large volume of peripherally related studies that do not meet the stringent inclusion criteria upon detailed screening.

Second, a substantial proportion of excluded records (n=622; 78% of screening exclusions) were studies that employed remote sensing in heritage contexts but did not substantively address climate-related threats or climate-adaptive monitoring frameworks. Many of these studies focused on documentation, visualization, or restoration planning without explicit consideration of climate change impacts, which is a core requirement for inclusion in this review.

Third, the review prioritizes methodological depth and application relevance over exhaustive coverage. Many excluded studies (n=89; 11% of screening exclusions) contained relevant technologies but lacked sufficient technical detail or empirical validation to contribute meaningfully to the synthesis of best practices.

Fourth, the four-decade temporal window (1990–2025) encompasses early-phase studies (1990–2000) where LiDAR and SAR applications in heritage were nascent and largely experimental. The majority of methodologically robust, climate-relevant studies are concentrated in the past 10–15 years, reflecting the maturation of these technologies and the growing urgency of climate-adaptive conservation.

Consequently, the 70 included studies represent the most technically substantive and directly relevant contributions to the specific research question, rather than an exhaustive inventory of all heritage remote sensing literature. This focused approach is consistent with PRISMA guidelines for systematic reviews that address specialized, technology-driven topics, where relevance and methodological quality are prioritized over sheer volume (Page et al., 2021). By adhering to PRISMA, this methodology ensures a robust, evidence-based synthesis, advancing the discourse on remote sensing-based climate-resilient heritage conservation. Theadvantages of the PRISMA technique are mentioned in Table 3.

Table 3. Advantages of PRISMA review over other review techniques

3.1. Preferred Journals

From the literature review, it has been observed that research on LiDAR and SAR applications for cultural heritage monitoring is prominently published in interdisciplinary journals bridging geospatial science, archaeology, and climate studies. Leading journals include Remote Sensing (MDPI), ISPRS Journal of Photogrammetry and Remote Sensing, Journal of Cultural Heritage (Elsevier), and Archaeological Prospection (Wiley), which emphasize technological innovations like InSAR deformation analysis and AI-driven LiDAR processing. Sustainability (MDPI) and Climate Risk Assessment (Elsevier) feature studies on climate-adaptive conservation, while IEEE Transactions on Geoscience and Remote Sensing dominates high-impact sensor methodology (Liu, 2024; Raimondi et al., 2009; Rodríguez-Gonzálvez et al., 2017; Trier et al., 2021; Vinci et al., 2025). The International Journal of Heritage Studies (Taylor and Francis) (Harvey, 2001; Winter, 2014), integrates policy perspectives and Heritage Science (Springer) (Pocobelli et al., 2018), focusing on multi-disciplinary case studies, particularly from the Global South.

The top journals by focus area are shown in Table 3. The dominance of interdisciplinary journals reflects the evolving nature of heritage monitoring, which increasingly integrates geospatial science, environmental modeling, and conservation practices. However, a noticeable gap exists in region-specific journals from the Global South, indicating an imbalance in knowledge production and dissemination. Table 4 summarises the top journals publishing in heritage conservation using remote sensing technology. The percentage share of the journals based on LiDAR and SAR remote sensing for heritage conservation is shown in Fig. 2. The percentage values represent the proportion of the 70 studies included in the final synthesis that were published in each journal. For studies appearing in multiple publications (e.g., dissertation chapters later published as articles), only the peer-reviewed journal version was counted.

Table 4. Top journals by focus area

Fig. 2. Research share by journals among the 70 studies included in the qualitative synthesis. Percentages are calculated as the number of included studies published in each journal divided by the total number of included studies (n=70).

4. 3-D LiDAR Mapping of the Dutch House (now Patna College Building)

To bridge the gap between theoretical developments and practical implementation, this review also incorporates representative case studies demonstrating the application of LiDAR and SAR techniques in real-world heritage monitoring scenarios. The Patna College, also known as the Dutch house, is a significant historical landmark in Patna city, Bihar, India. Constructed in the 18th century during the era of the Dutch East India Company, it served as a trading post and residence of the Dutch merchants (Kumawat, 2020). The architecture reflects the colonial style of the period, characterised by robust brickwork, large arches, and spacious verandas, representing a fusion of European design adapted to the Indian climate (Sinha, 2019).

With the decline of the Dutch colonial interests, the building was repurposed and became the administrative centre of the esteemed Patna College founded in 1863 (Brown, 2003). The transition of the building from a trading outpost to a centre of learning adds to its layered historical narrative (Sinha, 2019).

The 3-D LiDAR mapping of the Patna college building serves as a critical case study in the application of TLS for the documentation of the historically significant architecture, vulnerable to climate impacts. The methodology adopted the utilisation of a high-resolution TLS system- the TRIMBLE X9 scanner around the structure to capture the millions of precise geometric measurements, thus creating a dense millimetre-accurate point cloud. This raw data was then utilised to generate a digital twin, a high-fidelity 3-D model that accurately represents the building’s current state, including the intricate architectural details and any existing structural imperfections. This model now serves as a baseline reference for future multi-temporal change detection. While a repeat TLS survey is planned for 2027, the study demonstrates how the M3C2 algorithm (described in Section 4.1) can be applied once a second-epoch point cloud becomes available, enabling millimetre-level identification of deformation, crack propagation, or material loss. The overall methodology adopted for the 3-D LiDAR mapping of Patna College building is shown in Fig. 3. This case study illustrates the capability of terrestrial LiDAR systems to capture high-resolution structural details, enabling accurate documentation and condition assessment of heritage buildings at a local scale. The results demonstrate that LiDAR-based mapping provides highly reliable geometric measurements; however, its application remains constrained by logistical challenges, cost, and limited spatial coverage for large heritage complexes.

Fig. 3. Methodology for 3-D LiDAR model generation.

• Scanning Configuration and Data Acquisition

The terrestrial laser scanning survey of the Dutch House (Patna College) was conducted using a Trimble X9 3D laser scanner, a phase-based instrument with a stated range accuracy of ±2 mm and angular accuracy of 18″. A total of 17 scan positions were established around the building perimeter, with an average inter-scan distance of approximately 8 m to ensure adequate overlap (>30%) between adjacent scans. The scanner was mounted on a survey-grade tripod at a height of 1.6 m above ground at each position. Scanning parameters were set to a resolution of 6 mm at 10 m distance, yielding a mean point spacing of approximately 5–8 mm on the building façade. The total scan time per position averaged 4 min for a full-dome scan (360° × 300°).

• Registration and Point Cloud Processing

Scan registration was performed using Trimble RealWorks software, employing a combination of automated cloud-to-cloud registration (Iterative Closest Point algorithm) and manual target-based refinement using six spherical targets (diameter 145 mm) distributed across the site. The final registration achieved a mean absolute error of 3.2 mm across all overlapping regions, with a maximum reported error of 5.8 mm at one junction. The registered point cloud consisted of approximately 187 million points, which were then cleaned to remove vegetation, moving objects, and noise, resulting in a final dataset of 164 million points for model generation. The point cloud was subsequently decimated to a uniform 1 cm point spacing for 3D mesh generation and 2 cm for the M3C2-ready reference model.

4.1. The Multi-Scale Model to Model-Cloud Comparison (M3C2) Algorithm

Although M3C2 was not applied in the present study due to the lack of a repeat scan, the algorithm constitutes the recommended methodological framework for future monitoring of the Dutch House. The following description outlines its working principle and relevance to heritage conservation. The M3C2 is a sophisticated algorithm for the detection and quantification of change between two 3-D point clouds, such as those from LiDAR scans taken at different times (DiFrancesco et al., 2020). Unlike the simple methods that calculate direct distances between points, the M3C2 measures changes precisely along the local surface normal, thus making it highly precise and robust against noise and varying point densities (Zahs et al., 2022).

Crucially for heritage conservation, M3C2 enables the detection of millimetre-level deformations, such as crack propagation, subsidence, or erosion in structures like ancient forts or monuments (Qiao & Butt, 2023). This allows the conservators to monitor the structural health with high precision, assess vulnerabilities from climate stressors, and evaluate the effectiveness of preservation efforts over time (Seo et al., 2021).

5. SAR Remote Sensing-Based Displacement Analysis of Shergarh Fort in India

The Shergarh Fort located in Dholpur Rajasthan, India, is a historical citadel with a rich and layered past. While the region has ancient roots, the fort’s current structure is largely attributed to the Afghan ruler Sher Shah Suri, in the 16th century, who built upon the earlier foundations (The Islamic heritage, 2025). It served as a key strategic outpost controlling the vital trade and military routes between the Mughal heartland and the Deccan plateau. The fort complex encompasses sprawling battlements, palaces, water reservoirs and intricate underground structures, showcasing a blend of Hindu and Islamic architectural styles. Its strategic location on a rocky hilltop overlooking the Chambal River made it a prized possession, contested by the Mughals, Marathas and the British Empire. Today, it stands as a nationally protected monument of national importance, serving as a silent witness to centuries of Indian history. While the region has ancient roots, the fort’s current structure is largely attributed to the Afghan ruler Sher Shah Suri in the 16th century, who built upon earlier foundations (The Islamic heritage, 2025).

It served as a key strategic outpost controlling vital trade and military routes between the Mughal heartland and the Deccan plateau. The fort complex encompasses sprawling battlements, palaces, water reservoirs, and intricate underground structures, showcasing a blend of Hindu and Islamic architectural styles. Its strategic location on a rocky hilltop overlooking the Chambal River made it a prized possession, contested by the Mughals, Marathas, and the British Empire. Today, it stands as a nationally protected monument of national importance, serving as a silent witness to centuries of Indian history.

These site characteristics—particularly the extensive areal coverage (several hectares), the predominance of exposed rock and well-preserved masonry, and the abundance of man-made structures—make Shergarh Fort highly suitable for SAR-based monitoring using PSInSAR. The rocky hilltop provides stable, coherent scatterers over time, while the fort’s large spatial extent aligns well with the wide-area synoptic view offered by spaceborne SAR sensors such as Sentinel-1, enabling systematic deformation mapping without the need for ground instrumentation across the entire complex.

5.1. Methodology Adopted: Persistent Scatterer Interferometry (PSInSAR)

For the PSInSAR-based Line of Sight (LOS) surface displacement mapping and estimation of Shergarh Fort, a time series of Sentinel-1 SAR Single Look Complex (SLC) datasets was utilized. A total of 4 images were processed (since the time period selected is small and the aim was to showcase the technology for this review), comprising 2 from the ascending pass and 2 from the descending pass, spanning January 2022 to January 2026. All images were acquired with VV polarization. The detailed step-by-step methodology is illustrated in Fig. 4. It should be noted that the limited number of SAR acquisitions in this study constrains the robustness of time-series deformation analysis; therefore, the results should be interpreted as indicative rather than definitive.

Fig. 4. PSInSAR methodology flow diagram.

• Step 1: Data Pre-Processing (Split, Calibrate, Deburst)

The raw Sentinel-1 SLC data, acquired in Terrain Observation with Progressive Scans (TOPS) mode, were first split to isolate the swath covering the Shergarh Fort area. Radiometric calibration was applied to convert the digital pixel values to meaningful backscatter intensity (sigma naught). Subsequently, a deburst operation was performed to remove the gaps between the individual bursts, creating a continuous, single SLC image for each acquisition.

• Step 2: Co-Registration

All secondary SLC images were precisely co-registered to a single master image (typically the earliest acquisition). This process aligns the pixels of all images to sub-pixel accuracy, ensuring that each ground target contributes to the same pixel in every image of the stack. This is a critical step for reliable interferometric analysis.

• Step 3: Interferogram Generation

For every image pair (or for a network of pairs), an interferogram was generated by cross-multiplying the master SLC image with the complex conjugate of the co-registered slave SLC images. This process extracts the phase difference (interferometric phase) between the two acquisitions, which contains information related to ground displacement, topography, atmosphere, and noise.

• Step 4: Persistent Scatterer (PS) Point Selection

The core of the PSInSAR technique involves identifying pixels dominated by a single strong scatterer that remain coherent (stable) throughout the entire time series. These are known as PS points. Candidates were selected based on their temporal stability of backscatter, typically using a threshold on the Amplitude Dispersion Index (Here, Da=0.25, randomly selected). These points, often corresponding to man-made structures like the fort’s walls, battlements, and exposed rock, act as natural radar benchmarks. Candidates were selected based on their temporal stability of backscatter, using the standard amplitude dispersion index threshold of Da ≤ 0.25, as established by Ferretti et al. (2001) for persistent scatterer identification.

• Step 5: Sparse Time Geocoding and Phase Unwrapping

The stack of interferograms was used to perform a sparse time-series analysis. The phase of the selected PS points was analyzed to estimate and remove the topographic phase contribution using an external DEM. The remaining phase was then unwrapped to obtain a continuous phase value. This step resolves the ambiguity in the phase measurements and provides a cumulative phase change over time for each PS point.

• Step 6: Multi-Looking and Geocoding

While the analysis is performed on sparse PS points, this step in the diagram refers to the geocoding process, in which the radar coordinates of the results are transformed into a standard geographic coordinate system (e.g., latitude/longitude or UTM). This allows the displacement measurements to be overlaid on accurate maps and optical imagery for direct spatial correlation with the fort’s structures.

• Step 7: Phase to Line of Sight (LOS) Displacement Conversion

The final and crucial step involves converting the unwrapped phase values for each PS point into physical displacement along the satellite’s LOS. This conversion is governed by the fundamental InSAR equation 4. By utilizing both ascending and descending datasets, LOS displacements from different viewing geometries can be decomposed to estimate near-vertical and east-west components of motion. The current analysis is a workflow demonstration only. No independent ground-truth data (e.g., Global Navigation Satellite System, GNSS) surveys, tiltmeters, or crack gauges) were available for validation. Consequently, the displacement estimates produced here should not be considered verified measurements.

6. Results and Discussion

The results in this section focus on the measurements based on the LiDAR and PSInSAR techniques for mapping and monitoring the structural integrity and surface displacement of cultural heritage sites. Though the government agencies worldwide and in India are taking measures for conservation and restoration of the cultural heritage sites, yet with the changing climate, the regular mapping and monitoring are crucial as the climate-induced changes might not be immediate, many times but become evident over a period of many years, often causing irreparable damage.

6.1. Results

The terrestrial LiDAR survey of the Dutch House generated a high-resolution 3D point cloud with a mean point spacing of 6 mm on the primary façades. Dimensional validation was total station (Leica TS06, angular accuracy 2″). The checkpoints conducted using 24 independent check points measured with a were targeted at distinct architectural features (corners, window edges, cornices) and compared with their coordinates extracted from the LiDAR point cloud. The comparison yielded an Root Mean Squared Error (RMSE) of 1.3 cm, a mean absolute error of 1.1 cm, and a standard deviation of 0.7 cm. The maximum observed deviation was 2.4 cm at a partially occluded cornice feature. The overall accuracy of ±2 cm reported in this study refers to the 95% confidence interval (1.96 σ), confirming that the LiDAR model meets the sub-2-cm accuracy threshold required for heritage deformation monitoring. Table 5 summarizes the accuracy assessment statistics. The results for the Dutch house (now Patna College) are shown in Fig. 5.

Fig. 5. Rendered 3D digital twin of the Dutch House (Patna College) generated from terrestrial laser scanning.The model captures architectural details, including brickwork patterns, cornices, and window arches, providing a baseline for future M3C2 change detection analyses: (a) height estimation using a total station, (b) height estimation using a 3-D LiDAR model, (c) facade length measurement using a laser distometer and ground measurements, and (d) façade length measured using a LiDAR 3-D model.

Table 5. Accuracy assessment of the Dutch House TLS model against total station measurements

The accuracy assessment statistics of the LiDAR survey of the Dutch house are summarised in Table 5.

A similar LiDAR survey was carried out for the Indian Institute of Technology (IIT) Patna campus in Bihta, Patna district, Bihar, India, and the results are shown in Fig. 6.

Fig. 6. LiDAR survey of IIT Patna campus: (a) actual photograph of IIT Patna campus and (b) LiDAR 3-D model of IIT Patna campus.

The dimensions of the different sections of the buildings (both the Dutch house and the IIT Patna administrative building) were cross-checked as per the drawings and using a Total station and LiDAR readings, and a maximum difference of only cm was observed. The PSInSAR-based LOS displacement map of Shergarh fort is shown in Fig. 7.

Fig. 7. PSInSAR-derived displacement map of Shergarh Fort. (a) Google Earth image of Shergarh Fort with the boundary delineated in red. (b) Google Earth overlay of the PSInSAR-derived Line-of-Sight (LOS) displacement map (in millimetres). Colour convention: Positive values (red hues) indicate movement toward the satellite (surface uplift or structural outward displacement); negative values (blue hues) indicate movement away from the satellite (subsidence or inward displacement). The map is generated from a limited 4-image stack and should be interpreted as a methodological illustration (see Section 5 for limitations).

From Fig. 7, it is observed that the Shergarh Fort is experiencing a displacement of ±2 cm over a period of 4 years from January 2022 to January 2026. The limited image stack (n=4) resulted in a sparse and uneven distribution of persistent scatterer candidates. For a typical PSInSAR analysis with 20+ Sentinel-1 acquisitions, PS density in built-up heritage areas can exceed 100 points/km². With only 4 images, the number of stable scatterers identified was significantly lower, and their spatial distribution was insufficient to generate a reliable, spatially continuous deformation map. The displacement values shown in Fig. 7 should therefore be interpreted with extreme caution. A PS point density map is not presented here, as the sparse candidate distribution would be misleading rather than informative.

The reasons for the displacement are manifold and warrant detailed geotechnical and structural investigation. While climate-related factors—such as altered precipitation patterns or groundwater fluctuations—may contribute to such deformations, establishing a direct causal link requires further integrated analysis combining InSAR observations with environmental and geotechnical data. The area of Shergarh Fort is large, and the LiDAR equipment could not be carried to the actual site; therefore, PSInSAR-based analysis was only carried out. The area of Shergarh Fort is large, and the LiDAR equipment could not be carried to the actual site; therefore, PSInSAR-based analysis was only carried out. Since the Sentinel-1 SAR data has a spatial resolution of 5 m by 20 m and 14 m by 14 m upon multilooking with a swath of 5 km by 20 km, the Dutch house was too small a location for PSInSAR analysis. Both the LiDAR and PSInSAR are invaluable remote sensing techniques for cultural heritage mapping and structural integrity monitoring. However, the application depends upon many factors, such as the area to be covered and accessibility. In this study, field validation of the Shergarh Fort could not be carried out due to travel constraints, which is a limitation. Given the relatively small magnitude of displacement, uncertainties associated with atmospheric effects, temporal decorrelation, and noise inherent in InSAR processing should also be considered when interpreting these results.

From Fig. 7, a cumulative Line-of-Sight variation of ±2 cm was observed over the four-year period (January 2022 to January 2026). However, given the extremely limited SAR data stack (4 images), this value is well within the noise floor of the analysis. Sentinel-1 C-band interferograms are susceptible to atmospheric phase delays (ionospheric and tropospheric), orbital inaccuracies, and temporal decorrelation. With only four acquisitions, it is not possible to separate a genuine displacement signal from these noise sources. The ±2 cm variation should therefore not be interpreted as a measured deformation but rather as an illustration of the type of output that PSInSAR can generate when a sufficiently dense time series is available.

6.2. Discussion

The PSInSAR analysis of Shergarh Fort, utilizing Sentinel-1 time-series data from January 2022 to January 2026, revealed cumulative line-of-sight displacements of ±2 cm across the fort complex over the four-year observation period. This finding is particularly significant as it demonstrates the capability of spaceborne SAR interferometry to detect millimetre-scale surface deformations at heritage sites where ground-based instrumentation may be impractical due to topographical constraints, site inaccessibility, or the sheer spatial extent of the monument. The identification of ongoing displacement at Shergarh Fort raises important questions regarding the causative mechanisms, which may include a combination of geotechnical factors such as foundation settlement on the rocky hilltop, hydrological changes affecting slope stability in the Chambal River valley, and climate-induced stressors, including altered precipitation patterns and groundwater fluctuations (Neidhardt & Shao, 2023).

While detailed geotechnical investigations would be necessary to definitively attribute causality, the observed deformation patterns align with broader concerns regarding the vulnerability of heritage structures situated in climatically sensitive landscapes, where gradual environmental changes may manifest as progressive structural instability over decadal timescales. The finding is relevant primarily as a methodological proof-of-concept. The small image stack precludes any attribution of the observed variations to geotechnical, hydrological, or climatic processes. Establishing causal links between InSAR observations and heritage degradation will require (a) a multi-year stack of at least 30 Sentinel-1 acquisitions, (b) application of atmospheric phase correction, (c) integration of local rainfall, temperature, and groundwater level data, and (d) complementary structural surveys (e.g., TLS or crack gauges) for validation.

The complementary nature of LiDAR and SAR techniques becomes evident when considering the scale-dependent applicability of each technology. As observed in this study, the Dutch House’s relatively compact footprint rendered it unsuitable for PSInSAR analysis, given the spatial resolution limitations of Sentinel-1 data (5 × 20 m native resolution, approximately 14 × 14 m after multilooking), whereas the extensive Shergarh Fort complex covering several hectares provided sufficient persistent scatterer targets for meaningful interferometric analysis. Conversely, the millimetre-level detail captured by TLS for the Dutch House would be logistically challenging to replicate across the entire Shergarh Fort area given current equipment constraints and site accessibility issues. This scale-dependent complementarity suggests that optimal heritage monitoring strategies should adopt a tiered approach: spaceborne SAR for regional-scale screening and deformation hotspot identification, airborne or UAV-based LiDAR for intermediate-scale topographical mapping and vegetation penetration in forested archaeological landscapes, and terrestrial LiDAR or structured light scanning for high-detail documentation of individual structures and architectural elements requiring sub-centimetre precision. This scale-dependent complementarity forms the foundation for integrated, multi-tiered monitoring frameworks that can effectively address both local structural assessments and regional deformation dynamics. Some important studies are summarised in Table 6.

Table 6. Representative case studies: LiDAR and SAR applications in climate-exposed heritage sites

The recent literature synthesized in Table 4 reinforces the trajectory toward multi-sensor integration and algorithmic advancement in heritage monitoring applications. Said and Shahrin (2025) demonstrated the efficacy of LiDAR-photogrammetry fusion for comprehensive defect assessment, establishing standardized workflows that enhance documentation efficiency while maintaining the precision required for condition-based conservation decision-making. Pinheiro et al. (2024) contributed critical insights into the mechanisms linking climatic variables—temperature, humidity, and microclimatic fluctuations—to material degradation processes in historic buildings, providing an understanding of environmental processes that complements the structural observations derived from remote sensing. The multi-scale framework proposed by Fu et al. (2026), integrating satellite SAR for wide-area deformation monitoring with LiDAR and drone-based surveys for high-resolution structural assessment, represents a methodological advancement that directly addresses the scale-dependency challenges encountered in the present study. Their subsequent application of advanced multi-temporal InSAR techniques at the Summer Palace in Beijing (Tang et al., 2016) further validates the capability of persistent scatterer interferometry to detect subtle deformation signals (millimetre-scale annual displacement rates) at UNESCO World Heritage sites, while the comprehensive systematic review by Chen et al., (2023), situates these technological applications within the broader context of climate-adaptive heritage conservation, identifying emerging research directions and persistent gaps in the literature.

A critical observation emerging from this synthesis concerns the geographical distribution of advanced remote sensing applications in heritage conservation. Despite the demonstrated efficacy of LiDAR and SAR techniques, their adoption remains heavily concentrated in European, East Asian, and North American contexts, with the Global South—including South Asia, Africa, and parts of Latin America—remaining significantly underrepresented in the published literature. This disparity, highlighted by the 2024 UNESCO report noting that fewer than 5% of African heritage sites utilize advanced monitoring technologies, reflects the compound challenges of equipment costs (Eze et al., 2024), technical expertise requirements, data processing infrastructure limitations, and policy framework gaps that constrain technology transfer to resource-limited settings (Chen et al., 2023). The present study’s focus on Indian heritage sites—the Dutch House in Bihar and Shergarh Fort in Rajasthan—represents a modest contribution toward addressing this geographical imbalance, yet the challenges encountered in field validation for Shergarh Fort due to travel constraints underscore the practical difficulties of implementing comprehensive monitoring programs in regions where institutional support and research infrastructure may be limited.

6.2.1. Key Challenges and Research Directions

Despite significant advancements, several challenges persist in the application of LiDAR and SAR for cultural heritage monitoring. These include limited temporal resolution of LiDAR datasets, coherence loss in SAR data over vegetated or complex surfaces, and the absence of standardized workflows for integrating multi-sensor data. Furthermore, high operational costs, limited accessibility, and a lack of technical expertise restrict the adoption of these technologies in developing regions. Future research should focus on multi-sensor data fusion, machine-learning-based automation, and the development of scalable, cost-effective monitoring frameworks that address both structural and environmental dynamics.

6.2.2. Possible Policy Implications

The implications of these findings for heritage conservation policy and practice are manifold. First, the demonstrated capability of PSInSAR to detect displacement at Shergarh Fort suggests that spaceborne remote sensing should be integrated into national heritage monitoring frameworks as a routine screening tool, enabling archaeological survey organizations and monument conservation authorities to identify emerging stability issues before they manifest as visible structural damage requiring costly emergency interventions. Second, the establishment of baseline 3-D documentation through TLS, as demonstrated for the Dutch House, should be prioritized for heritage assets in climate-vulnerable locations, providing the reference data against which future change-detection algorithms—such as M3C2—can quantify deformation rates and inform condition assessments. Third, the growing body of literature demonstrating successful integration of machine learning techniques with LiDAR and SAR data points toward the need for capacity-building initiatives that train heritage professionals in geospatial data analytics, ensuring that technological capabilities are matched by human expertise. Fourth, the emergence of low-cost SAR microsatellite constellations (e.g., ICEYE, Capella Space) and increasingly accessible UAV-based LiDAR systems offers pathways toward democratizing access to these technologies, though policy mechanisms will be necessary to ensure that cost reductions translate into equitable access for heritage institutions across the Global South. These findings emphasize the need for policy frameworks that integrate remote sensing technologies into routine heritage monitoring programs, particularly in climate-vulnerable regions.

6.2.3. Limitations

The limitations encountered in this study merit explicit acknowledgment. The inability to conduct field validation for the PSInSAR displacement measurements at Shergarh Fort, while necessitated by practical travel constraints, limits the certainty with which the observed deformation signals can be attributed to specific structural elements or geological processes. Future research should prioritize integrated field campaigns that combine spaceborne SAR analysis with ground-based instrumentation—including GNSS surveys, tiltmeters, or crack gauges—to validate satellite-derived displacement measurements and establish empirical relationships between InSAR observations and physical structural responses. Additionally, the temporal coverage of the PSInSAR analysis (four years) provides an initial indication of displacement trends but may not capture longer-term cyclical patterns or acceleration phenomena that could signal impending instability. Continued monitoring over extended timeframes, coupled with the incorporation of historical SAR archives (e.g., ERS-1/2, ENVISAT ASAR), would enable more robust trend analysis and improved understanding of deformation dynamics in relation to climatic variables. Addressing these limitations will require integrated field validation, longer temporal datasets, and improved methodological standardization across studies.

The PSInSAR analysis of Shergarh Fort is subject to significant limitations. The image stack (4 images) is far below the recommended minimum for PSInSAR (15–30 acquisitions), the Amplitude Dispersion Index threshold was applied as per standard practice rather than empirically validated for this site, PS point density was insufficient to produce a robust deformation map, and the observed ±2 cm variation lies within the expected noise envelope. Furthermore, no field-based validation (e.g., GNSS, geotechnical instrumentation) was possible due to travel constraints. These results should therefore be viewed solely as an illustration of the PSInSAR workflow and its potential for future monitoring, not as a scientific measurement of deformation at Shergarh Fort.

6.3. Proposed Integrated LiDAR–SAR Monitoring Framework for Climate Adaptive Heritage Conservation

A central contribution of this review is the conceptualization of a tiered, multi sensor framework that combines LiDAR’s spatial precision with SAR’s temporal regularity to address both structural documentation and dynamic deformation monitoring at heritage sites. While the two case studies in Sections 4 and 5 demonstrated the independent application of Terrestrial Laser Scanning (Dutch House) and PSInSAR (Shergarh Fort), these were isolated demonstrations constrained by site specific logistics. The integrated framework presented here (Fig. 8) proposes a systematic methodology that overcomes these isolation constraints by merging LiDAR derived baseline models with SAR based time series deformation analysis within a common geospatial framework. The framework is structured into four operational tiers, each corresponding to a distinct spatial scale and monitoring objective, and is designed to be adaptable to site size, accessibility, and available resources. The Proposed Integrated LiDAR–SAR Monitoring Framework for Climate Adaptive Heritage Conservation is shown in Fig. 8.

Fig. 8. Proposed integrated LiDAR-SAR framework for climate adaptive heritage conservation monitoring.

Tier 1 – Regional Screening (Spaceborne SAR): Wide area deformation screening using Sentinel 1 C band SAR data acquired with a minimum temporal stack of 30 images (ideally spanning >3 years). PSInSAR and SBAS InSAR algorithms are applied to identify deformation hotspots and zones of anomalous subsidence or uplift. Ancillary datasets (precipitation, groundwater levels, temperature) are ingested to contextualize displacement signals.

Tier 2 – Site Scale Topographic Mapping (UAV or Airborne LiDAR): Once deformation hotspots are identified, high resolution topographic mapping is conducted using UAV borne LiDAR (or airborne LiDAR for extensive complexes) to generate a DEM and a coarse structural model. This tier provides the topographic reference for correcting InSAR phase contributions and identifies areas requiring detailed structural documentation.

Tier 3 – High Detail Structural Documentation (Terrestrial LiDAR): Terrestrial Laser Scanning is deployed at the monument or building scale to capture sub centimetre architectural detail. Multiple scan positions are registered with survey grade accuracy, producing a dense point cloud that serves as the baseline 3D digital twin. This baseline is archived for future multi temporal change detection using algorithms such as M3C2.

Tier 4 – Integrated Analysis and Decision Support: The outputs of Tiers 1–3 are fused in a GIS environment. SAR derived displacement time series are projected onto the LiDAR derived 3D model to correlate deformation patterns with specific structural elements. Machine learning classifiers trained on multi temporal LiDAR difference maps and InSAR velocity fields detect anomalous change. The system generates two primary outputs: (1) a structural health index that quantifies degradation severity per architectural element, and (2) a risk priority map that guides conservation intervention. A feedback loop enables adaptive re survey scheduling: if deformation rates exceed a predefined threshold (e.g., 2 mm yr–1), the system triggers an unscheduled TLS or UAV survey to capture the change with high spatial resolution.

7. Conclusions

This PRISMA-based review systematically charted the evolution of LiDAR and SAR remote sensing for climate-adaptive heritage monitoring, from early experimental applications to contemporary multi-sensor frameworks. A key contribution is the formalization of a tiered integrated LiDAR–SAR monitoring framework, which operationalizes the complementarity between LiDAR’s sub-centimetre structural documentation and SAR’s millimetre-scale deformation tracking. While separate components of this framework were demonstrated through the Dutch House LiDAR survey (Tier 3) and the Shergarh Fort PSInSAR workflow (Tier 1), full integration across all tiers remains a priority for future field campaigns. The review also identified persistent geographical imbalances, with the Global South significantly underrepresented in published integrated monitoring studies.

This review proposes a tiered, integrated LiDAR–SAR monitoring framework that, while demonstrated here only through separate representative components, provides a structured pathway toward unified multi-sensor heritage monitoring. Its full implementation awaits collaborative field campaigns that combine SAR time series, high-resolution LiDAR, and ground-truth validation. The two case studies included in the manuscript—the 3D LiDAR mapping of the Dutch House (Section 4) and the PSInSAR processing of Shergarh Fort (Section 5)—illustrate individual components of the framework but do not constitute a fully integrated implementation. Specifically, the Dutch House study demonstrated Tier 3 capability (baseline documentation) but lacked SAR-based deformation data, while the Shergarh Fort study demonstrated Tier 1 potential but was limited by an insufficient image stack and the absence of LiDAR-based structural correlation. The full integration, including co-registration of deformation measurements with a LiDAR-derived digital twin and automated risk assessment, has not yet been executed on a single heritage site and represents a clear direction for future research. This distinction is critical to avoid overstating the immediate applicability of the proposed framework and to outline the necessary next steps for operationalization clearly.

The future of climate-adaptive heritage monitoring lies in the convergence of multi-sensor remote sensing, advanced analytics, and scalable implementation frameworks capable of addressing complex and evolving environmental challenges. The integration of LiDAR, SAR, optical imagery, hyperspectral data, and in situ sensor networks within unified data fusion architectures will enable comprehensive characterization of structural integrity, surface processes, and deformation dynamics across multiple spatial and temporal scales. Such synergistic approaches move beyond isolated observations toward holistic monitoring systems that can capture both gradual degradation and abrupt damage mechanisms associated with climate-induced stressors.

The increasing availability of long-term satellite archives, combined with the emergence of high-frequency SAR constellations and UAV-based LiDAR platforms, provides unprecedented opportunities for continuous monitoring, near-real-time assessment, and early warning of structural instability at heritage sites. In parallel, advances in machine learning and artificial intelligence—including deep learning and explainable AI—are expected to significantly enhance automated feature extraction, anomaly detection, and predictive modelling of climate-driven risks. These developments will facilitate the transition from descriptive mapping toward predictive and prescriptive conservation strategies.

However, realizing these capabilities requires addressing several persistent challenges, including data accessibility constraints, a lack of standardized multi-sensor integration frameworks, and limited validation through field-based measurements. The underrepresentation of heritage sites in the Global South further highlights disparities in technological adoption, emphasizing the need for equitable research collaborations, capacity-building initiatives, and the development of cost-effective monitoring solutions. Open data ecosystems, cloud-based processing platforms, and low-cost satellite technologies will play a critical role in democratizing access and enabling wider implementation.

Furthermore, translating technological advancements into practical conservation outcomes necessitates stronger integration of remote sensing outputs into policy frameworks, heritage management practices, and climate adaptation strategies. Establishing standardized protocols, decision-support systems, and interdisciplinary collaboration between remote sensing scientists, conservation practitioners, and policymakers will be essential to ensure that data-driven insights effectively inform preservation efforts.

Collectively, these directions underscore a paradigm shift toward proactive, data-driven, and climate-resilient heritage conservation. By bridging technological innovation with practical implementation and policy integration, remote sensing will play a transformative role in safeguarding cultural heritage for future generations under the accelerating pressures of climate change. Ultimately, the integration of multi-scale remote sensing with climate-informed decision frameworks will redefine the future of heritage monitoring and conservation.

Author Contributions

Conceptualization, Methodology, Investigation, Data curation, Formal analysis, Visualization, Writing–original draft: Tripathi A; Datacuration, Investigation, Resources: Kumar P; Conceptualization, Funding acquisition, Supervision, Project administration: Pradhan B; Writing–review & editing: All authors.

Conflicts of Interest

No potential conflict of interest relevant to this article was reported. This article was commissioned by the Editorial Board and underwent independent peer review.

Funding

None.

Data Availability Statement

The data would be made available by the corresponding author upon reasonable request.