Introducation

Climate emergencies threaten the life of the planet and our current civilization. They endanger both cultural and natural heritage, while impacting the lives and well-being of populations worldwide1. The UNESCO Convention on the Protection of World Cultural and Natural Heritage requires the safeguarding of sites of outstanding universal value. The conservation of these locations requires consistent and stable environmental conditions2. In recent decades, China has experienced rapid economic growth and increased urbanization. The urbanization rate in China has reached 66.16%3,4,5,6. The proliferation of impermeable surfaces and significant changes in land cover have intensified the UHI effect throughout this developing phase7,8,9. The intensification of the UHI effect leads to increased temperatures in urban areas and their surroundings. The continual presence of environmental disruptions can cause changes in cultural legacy, often leading to related and enduring consequences10,11,12. Thus, understanding the impact of fluctuations in UHI effect intensity on the thermal conditions of cultural sites in urban peripheries is essential for heritage conservation.

Cultural heritage conservation requires precise microclimate regulation at preservation sites. Empirical evidence indicates that suboptimal illuminance accelerates material degradation in archaeological contexts, where prolonged solar exposure induces salt migration and recrystallization within structural matrices12,13. Poor ventilation in the heritage site area further promotes fungal growth, which may cause irreversible damage to cultural relics14. Among heritage typologies, cave temples represent particularly vulnerable systems. Constructed through subtractive lithic techniques or using porous earthen materials, these architectures inherently exhibit weathering susceptibility. Their mineralogical porosity mediates dynamic interactions with environmental parameters (thermal fluxes, hydric cycles, and biological activity), creating complex degradation pathways15,16. Over decadal timescales, synergistic effects of natural weathering and anthropogenic pressures have manifested as exfoliation, pigment loss, and structural compromise in numerous cave temple systems17.

Contemporary conservation research focuses on three primary domains: hydraulic infiltration processes, innovative tracer methods, and hydrological modeling clarify leakage paths and moisture transport dynamics18,19.Weathering kinetics and computational fluid dynamics simulations rebuild aeolian patterns, whereas chemo-mechanical models measure surface recession rates19,20. Hygrothermal behavior and coupled heat-moisture transport models delineate microenvironmental variations, illustrating how cyclic swelling and shrinkage pressures produce microfracture networks that extend into macroscale discontinuities21,22. These structural imperfections later serve as favored conduits for water penetration and expedited deterioration. While current research thoroughly details site-specific thermal-hydric regimes, significant information gaps remain concerning the macro-scale climatic influence of urban thermal plumes. The possible relationship between UHI amplification and the alteration of microclimates at cultural sites is poorly examined, especially about how metropolitan thermal anomalies can disrupt the fragile energy balance of nearby cave temple systems.

UHI phenomenon manifests as elevated temperatures in metropolitan regions compared to surrounding rural zones23,24. This thermal differential may extend to influence heritage sites situated on urban peripheries. The magnitude of temperature disparity is governed by climate change25,26,27, land use configurations28,29,30, urban thermal regimes31,32,33, and anthropogenic drivers34,35,36,37.Two primary typologies emerge based on measurement methodology: those quantified through land surface temperature (LST) obtained via remote sensing22,38,39, and those assessed via atmospheric temperature measurements. The present study focuses on the former category, specifically surface-level heat island effects23,40. Pioneering work by Rao established the application of thermal remote sensing for analyzing spatial heat patterns and their correlations with urban surface characteristics, marking the first satellite-based documentation of surface heat anomalies41. Subsequent methodological developments have significantly enhanced monitoring capabilities through diversified sensor-platform integrations.

Consequently, infrared remote sensing has become a pivotal methodology for characterizing surface thermal anomalies in urban systems, driving rapid advancements in observational techniques42,43.Quantifying thermal differentials between urban cores and peripheral zones remains a central focus in such investigations, particularly through the lens of heat accumulation intensity metrics44,45,46. Empirical analyses consistently reveal three fundamental spatial characteristics: 1) thermal intensity demonstrates exponential decay with increasing distance from urban centers39; 2) transect-based approaches effectively capture both the magnitude and spatial configuration of surface thermal gradients; 3) aggregated surface temperatures exhibit strong positive correlations with impervious surface coverage and negative correlations with vegetation indices34,40,47. Notably, orographic modulation in mountainous urban areas has been shown to reduce absolute temperatures without diminishing relative urban-rural thermal contrasts, suggesting topographic complexity does not fundamentally alter heat island dynamics47.

Currently, research on the thermal characteristics of cultural monuments and urban heat islands is fragmented. Research on the microenvironment of heritage sites mostly focuses on the site itself, addressing concerns such as leakage, weathering, salinization, and environmental monitoring technologies. Investigations on urban heat islands predominantly concentrate on metropolitan areas, namely the magnitude and severity of heat islands. Insufficient research exists about the impact of urban expansion on the thermal environment of cultural sites situated on the urban periphery. This study investigates the thermal conditions of heritage sites and evaluates the effects of urban microenvironmental changes on their microenvironments. This research presents three primary contributions: The impact of urban expansion on the cold island phenomenon of rivers in culturally significant areas. The impact of urban heat islands on the thermal intensity of impermeable surfaces in cultural site areas. The outcomes of this study seek to clarify ways for mitigating thermal changes in heritage sites caused by UHI effects, therefore improving the conservation of these sites on the urban periphery.

Methods

Framework

This study examines the influence of the UHI effect on the thermal environment of regions adjacent to cultural heritage sites on the urban periphery, employing Landsat satellite data and atmospheric correction techniques to conduct temperature inversion for both the urban area of Luoyang and the Longmen Grottoes site. The study examines the quantitative correlation between alterations in land cover types at the Longmen Grottoes site and the metropolitan region of Luoyang, utilizing China’s 30-m resolution land cover data from 1985 to 2023. Figure 1 illustrates the technological framework of our study. The framework comprises three components and encompasses four modules.

Fig. 1
Fig. 1
Full size image

Technical framework of this study.

PART 1: On-site measurements were done to examine the thermal environment of the Longmen Grottoes site area. Data monitoring was conducted in both winter and summer. Both fixed and mobile measurement techniques were employed for field testing. The fixed-point test results were utilized to calibrate the mobile test data, thereby removing temporal inaccuracies in the mobile measurements. The on-site measurements of the Longmen Grottoes area yield critical insights into the micro-scale temperature distribution, which is essential for comprehending the internal thermal environment of the site.

PART 2: Urban expansion results in alterations to the land surface temperature of Luoyang’s urban area. Temperature inversion data was utilized to acquire land surface temperature data for Luoyang’s urban region over the previous two decades. The integration of this data with various temporal dimensions over the past two decades is crucial for analyzing trends in land surface temperature variations in the urban region of Luoyang.

PART 3: Urban growth may adversely affect the thermal environment of cultural sites by altering UHI intensity. The urban-rural gradient analysis method was employed to get heat island intensity data for the Longmen Grottoes site area over the previous two decades. This data facilitated an analysis of the cold island effect of the Yihe River within the heritage site region, so broadening the research scope concerning the thermal environment of heritage sites on the urban periphery at a macro level.

PART 4: The linear correlation between urban built-up regions and heat island intensity in the Longmen Grottoes site area was examined using land cover data. This study seeks to examine the influence of urban growth on the environmental conditions of cultural sites located on the urban periphery. This research innovatively examines the thermal climate of cultural monuments from a macro perspective of urban development. It offers a novel viewpoint on environmental conservation in heritage sites located on the urban periphery. Integrating studies of local urban climates with investigations of the internal environments of cultural monuments enables a more precise evaluation of environmental alterations in areas surrounding these sites.

Analysis methods

Field measurement

The research involved data monitoring of the Longmen Grottoes site between winter and summer. A hybrid approach combining stationary and mobile measurement techniques was employed for field testing. The testing intervals occurred from January 23, 2024, to January 24, 2024, and from July 23, 2024, to July 24, 2024. The testing zone encompassed the western mountain visitor walkway of the Longmen Grottoes scenic region.

Mobile land surface temperature measurements were acquired using an infrared thermometer.Given the extensive expanse of the Longmen Grottoes Western Mountain visitor walkway, which surpasses the measuring capacity of the infrared thermal imager, a segmented mobile measurement methodology was employed. Three distinct photos were chosen from both winter and summer for first comparison examination. Figure 2 illustrates the locations of temperature measurements throughout the visiting walkway of the Longmen Grottoes Western Mountain. Surface temperature measurements at three test sites along the Western Mountain pathway were acquired using an infrared device, leading to the development of a surface temperature distribution map. A thermocouple was utilized to measure the surface temperature at specific locations due to the time delay linked with segmented measurements.The data from the stationary test locations was subsequently utilized to calibrate the mobile test data, thereby rectifying the temporal discrepancies in the mobile measurements. Fixed-point surface temperature measurements were performed with the multifunctional tester, which documented ground temperature, heat flux density, wind speed, and humidity data at designated locations at the southern entrance of the Longmen Grottoes Western Mountain scenic area. The data was recorded at a frequency of one point every five minutes. The designated temperature measurement point (a) is situated near the southern entrance of the Longmen Grottoes Western Mountain, where a mobile meteorological station has been established for air temperature assessments.

Fig. 2: Violin plot of mobile measurement data.
Fig. 2: Violin plot of mobile measurement data.
Full size image

a Winter test. b Summer test.

Temperature inversion

Presently, there exist three primary algorithms for surface temperature inversion: the atmospheric correction method, the single-window approach, and the split-window methodology. The atmospheric correction method use synchronized real-time atmospheric data from satellites to compute the attenuation equation of surface radiation during atmospheric transmission. It subsequently subtracts the attenuated radiation during transmission from the overall thermal radiation, resulting in a more precise measurement of surface thermal radiation intensity. The surface temperature is ultimately determined using a formula. This approach is the most prevalent for temperature inversion and yields the most precise results, with an inversion temperature error of under 0.6 K. The faults in the other two approaches are marginally greater. Consequently, this study employs the atmospheric correction method.

Radiometric calibration and atmospheric correction were first applied to the chosen satellite images. The processed data originated from calibrated and clipped outputs. The first retrieval of land surface temperature (LST) was performed utilizing the atmospheric correction approach. Extensive research by prior researchers has demonstrated that the essential mechanism of this approach functions as follows:

Step 1: Determine the surface emissivity (\(\varepsilon\)). The NDVI threshold approach, as described by Sobrino and other researchers, is employed for calculations based on the normalized difference vegetation index (NDVI) and fractional vegetation cover (FVC) derived from satellite data48.

$${FVC}=\frac{{NDVI}-{{NDVI}}_{{soil}}}{{{NDVI}}_{{veg}}-{{NDVI}}_{{soil}}}$$
(1)
$$\varepsilon =0.004{FVC}+0.986$$
(2)

In Eq. (1), FVC denotes the fraction of vegetation cover; NDVI signifies the Normalized Difference Vegetation Index; \({{\rm{NDVI}}}_{{\rm{soil}}}\) refers to the NDVI for regions devoid of vegetation cover; \({{\rm{NDVI}}}_{{\rm{veg}}}\) pertains to the NDVI for regions with complete vegetation cover; \(\varepsilon\) denotes the land surface emissivity(Eq. (2)) for the Landsat 7 ETM and Landsat 8 OLI_TIRS satellite sensors.

Step 2: Correction of atmospheric radiation and computation of thermal infrared radiance values. The atmospheric correction parameters for Landsat sensors are accessible on NASA’s official website (https://atmcorr.gsfc.nasa.gov). It can be calculated using the Eq. (3):

$${L}_{r}=\varepsilon \times \tau \times B\left({T}_{S}\right)+\left(1-\varepsilon \right)\times \tau \times {L}_{D}\downarrow +{L}_{u}\uparrow$$
(3)

Where \(\tau\) is the atmospheric transmittance, \({L}_{u}\uparrow\) and \({L}_{D}\downarrow\) represent the upward atmospheric radiance and downward atmospheric radiance, respectively. With τ = 0.66, \({L}_{u}\uparrow\) = 2.83,\(\,{L}_{D}\downarrow\) = 4.44 in this study.

The computation of blackbody radiance \(B\left({T}_{S}\right)\) can be articulated using the subsequent formula:

$$B\left({T}_{S}\right)=\frac{{L}_{r}-{L}_{u}\uparrow -\left. (1-{\rm{\varepsilon }}\right)\times \tau \times {L}_{D}\downarrow }{\varepsilon \times \tau }$$
(4)

Step 3: The intensity of blackbody radiation in Eq. (4) is transformed into land surface temperature in Eq. (5), utilizing a computation formula where \({K}_{1}\) and \({K}_{2}\) are predetermined values. The \({K}_{1}\) and \({K}_{2}\) values of Landsat 7 ETM are 666.09 and 1282.71, respectively, while Landsat 8 OLI_TIRS is 774.89 and 1321.08, respectively.

$${T}_{S}=\frac{{K}_{2}}{\mathrm{ln}\left (\frac{{K}_{1}}{B\left({T}_{S}\right)}+1\right)}$$
(5)

Upon completion of the retrieval, the resultant data underwent visualization processing, subsequently followed by the gridding of the land surface temperature (LST) data.This methodology is founded on the research conducted by Soe W. Myint et al. 49, with the grid dimensions established at 210 m × 210 m. A raster-to-point operation is executed to transform the raster image into vector point data suitable for analysis.The study area is partitioned into 18,700 equal-area grids, each measuring 210 m × 210 m. The surface temperature vector data is ultimately allocated to the grids, with the average temperature of the points within each grid serving as the grid’s temperature value.

Urban-rural gradient UHI

A geographical profile centered on the Longmen Grottoes research region is initially created based on the obtained land surface temperatures. This profile commences with the early city core of Xigong District in northern Luoyang, traversing the southern urban expanse of Luolong District, the southern suburban area of the Longmen Grottoes, and the wooded region of Wanan Mountain to the south. The reference zero point for the profile is located at the center of Xigong District in Luoyang, designated as point 0, while the endpoint is positioned in the western region of Wanan Mountain, encompassing a total of 632 points. The surface temperature data is allocated to the profile at intervals of 50 m, producing temperature distribution maps for the years 2001 and 2024, extending from the urban region to the Longmen Grottoes location, as depicted in Fig. 3b.

Fig. 3: Location of the research object: sitemap and sample areas of the Longmen Grottoes archaeological area.
Fig. 3: Location of the research object: sitemap and sample areas of the Longmen Grottoes archaeological area.
Full size image

a Geographical location of the study (China map sourced from http://bzdt.ch.mnr.gov.cn/). b Urban map of Luoyang City. c Aerial photograph of the Longmen Grottoes. d Details of the Grottoes.

The severity of the UHI effect can be quantified by the temperature differential between locations, with the extent of the heat island impact indicated by temperature variances across diverse areas50. This study utilizes the notion of inter-regional comparison to examine the variation in UHI intensity, focusing on the average land surface temperature differential between urban and rural areas. This study defines the rural area as the southern mountainous section of Wanan Mountain, characterized by vegetation covering exceeding 90%46. The average land surface temperature of the Wanan Mountain region was determined based on the temperature profile. The temperature differential, in relation to the urban area temperature, was utilized to ascertain the specific heat island intensity.Ultimately, for visualization, we constructed an area map illustrating the multi-temporal land surface temperature and UHI intensity profiles of the study area.

Data and materials

Test instruments

The mobile land surface temperature measurements were performed utilizing the Testo 890 infrared thermometer. Fixed-point surface temperature measurements were conducted with the Jiantong 2020 multifunctional tester. The specifications of the testing instruments are presented in Table 1.

Table 1 Measurement instrument parameter data
Full size table

Remote sensing satellite data

This study utilizes Landsat series satellites as the source of remote sensing data, with information acquired from the United States Geological Survey’s Earth Explorer website (https://earthexplorer.usgs.gov). This study utilizes vector data comprising administrative boundary files for the city and county levels of Luoyang, obtained from the Institute of Resources and Environment Science & Data Center of the Chinese Academy of Sciences (http://www.resdc.cn). Atmospheric profile parameters were acquired from the NASA Atmospheric Correction Parameter Calculator website (https://atmcorr.gsfc.nasa.gov, accessed before 2024). The land surface temperature data utilized for analysis after 2013 is derived from the temperature-retrieved land surface temperature product within the Landsat Collection 2 dataset, accessible via the USGS Earth Explorer website (https://earthexplorer.usgs.gov).

This study concentrates on data collected in July near Luoyang as the research subject. The data was obtained from the USGS website, specifically for strip number 125 and row number 036, which entirely encompasses the research region. The analysis encompasses cloud-free or low-cloud data exhibiting high remote sensing image quality from July throughout the previous two decades within the study area. The satellite remote sensing data is predominantly from July, with imaging timings focused around 3:00 AM GMT, ensuring a consistent temporal framework. The acquired land surface temperature data is thus both referenceable and comparable. Table 2 presents the satellite remote sensing data information.

Table 2 Satellite remote sensing data information
Full size table

Case description

Luoyang is situated in the central-eastern region of China, traversed by two rivers flowing east-west: the Luo River to the north and the Yi River to the south. The Longmen Grottoes are situated on both banks of the Yi River, with the primary grotto clusters lying on the western side of the river’s western peak. In the last two decades, Luoyang has undergone swift urbanization, as illustrated in Fig. 3b. In 2001, the primary urban zone of Luoyang was situated on the northern bank of the Luo River, with a compact urban built-up area around 13 km from the Longmen Grottoes location Fig. 4a. By 2016, Luoyang had nearly finalized the development of the Luolong District New Area, situated south of the Luo River, resulting in a substantial rise in impervious surfaces in the southern region of the Luo River. Conversely, the southern region of the Yi River had only intermittent impermeable surfaces resulting from the development of the Yi Bin New Area Fig. 4b. By 2024, impermeable surfaces in the southern region of the Luo River had markedly expanded, and the Yi Bin New Area adjacent to the southern section of the Yi River had undergone development, exhibiting significant expansion in impervious surfaces. The heavily constructed urban region is now roughly 4 km distant from the Longmen Grottoes site Fig. 4d. The Wanan Mountain region, situated in the southeastern sector of Luoyang, is characterized by rugged terrain primarily enveloped in woods. In the last two decades, this area has remained mostly unaltered and continues to be a forested region Fig. 3b. The urban configuration of Luoyang has transitioned from a linear city to a centralized urban region51.

Fig. 4: Land surface temperature distribution maps of Luoyang City for the years.
Fig. 4: Land surface temperature distribution maps of Luoyang City for the years.
Full size image

a–d Satellite image of Luoyang urban area in 2001, 2016, 2019 and 2024 respectively. e–h Land surface temperature of the urban area of Luoyang urban area in 2001, 2016, 2019 and 2024 respectively.

The Longmen Grottoes, a renowned UNESCO World Heritage site, comprise more than 2300 caves and recesses along a 1-kilometer-long sheer limestone cliff. The excavation of these caverns commenced in the late Northern Wei dynasty (late 5th century) and persisted throughout the early Tang Dynasty (mid-8th century). These caves and recesses include around 110,000 Buddhist statues, more than 60 Buddhist pagodas, and 2800 inscriptions. These sculptures represent a valuable treasure of stone artistry globally, encapsulating the wisdom and ingenuity of ancient civilizations. They possess considerable scientific significance and serve as a crucial reference in the examination of global cave art38. The Longmen Grottoes constitute a prominent large-scale semi-open cave temple complex hewn from limestone, directly influenced by sunlight, precipitation, and atmospheric pollution, resulting in an accelerated deterioration rate. Researchers have indicated that fluctuations in humidity and temperature within caves frequently influence the modification of soluble salts or induce intricate salt damage to the cave formations22. The Longmen Grottoes site has received emergency protection measures and is currently in the preventative conservation phase. The urban growth of Luoyang is expected to impact the thermal environment of the Longmen Grottoes site, rendering it a subject worthy of investigation from a preventive conservation standpoint.

Results

Analysis of measured surface temperature

The fixed ground temperature test point (a) is situated near the southern entrance of the western mountain of the Longmen Grottoes Scenic Area in Luoyang Fig. 5. The test outcomes for winter and summer are illustrated in Fig. 6a, b, respectively. The temperature disparity between the air and ground at the designated test location remains within 2 °C during both winter and summer, exhibiting minimal temperature fluctuation. The precise ground temperature data from the stationary site offers reliable calibration data for the mobile ground temperature observations.

Fig. 5
Fig. 5
Full size image

Map of mobile temperature measurement points.

Fig. 6: Temperature and air temperature measurements at fixed testing points.
Fig. 6: Temperature and air temperature measurements at fixed testing points.
Full size image

a Winter map. b Summer map.

Figure 5 illustrates the movable ground temperature test points. During winter, the infrared thermometer was employed to assess the temperature at various intervals, as detailed below: Test point P1 was recorded at approximately 16:00 on January 23, 2024, whilst test points P2 and P3 were recorded at 17:00. At 16:00, the ground temperature was noted as 0.0 °C, and at 17:00, it was −1 °C. The data recorded at 16:00 served as the calibration baseline, and the temperatures at test points P2 and P3 were modified by +1 °C. The median ground temperature readings for test points P1, P2, and P3 were 5.3 °C, 4.1 °C, and 6.6 °C, respectively Fig. 2a Measured winter surface temperatures along the western mountain tourist route at the Longmen Grottoes revealed the following patterns: The median ground temperature at test point P1 was 5.3 °C, while the ground temperature at test point P3 exceeded that of test point P2 by 2.5 °C. The surface temperature exhibited a gradient of elevated temperatures in the north and diminished temperatures in the south, possibly driven by the mountain’s orientation, which affects the duration of sunlight exposure to the surface.

During the summer, the infrared thermometer was utilized for segmented measurements as outlined below: Test point P1 was recorded at approximately 17:00 on July 23, 2024, whereas test points P2 and P3 were recorded at 17:30. A thermocouple indicated that the ground temperature was 33.7 °C at 17:00 and stayed at 33.7 °C at 17:30. Utilizing the 17:00 ground temperature data as the calibration baseline, the temperatures at test locations P2 and P3 remained constant. A map showing the distribution of surface temperatures was generated from field measurements, and a violin plot illustrating the surface temperature in the western mountain passage of the Longmen Grottoes was constructed. The median ground temperature readings for test points P1, P2, and P3 were 41.8° C, 41.5° C, and 38.5° C, respectively Fig. 2b.

Field measurements of the summer surface temperature along the western mountain tourist route of the Longmen Grottoes revealed the following patterns: The median ground temperature at test point P1 was 41.8 °C, whereas at test point P3 it was 38.5 °C, demonstrating that the ground temperature at P3 was 3.3 °C lower than at P1. The ground temperature in the western mountain path of the Longmen Grottoes progressively declines from south to north during the summer. This pattern is attributed to two factors: firstly, the general orientation of the mountain inside the heritage area; and secondly, the reduced vegetation coverage in the southern section of the Longmen Grottoes site relative to the northern section. An appropriate plant design in the Longmen Grottoes heritage area may enhance vegetation coverage and thus optimize the site’s thermal climate. Consequently, analyzing the temperature distribution patterns at the Longmen Grottoes site using field measurements is of considerable significance.

Analysis of temperature inversion

Surface temperature distribution maps for the metropolitan region of Luoyang in 2001, 2016, 2019, and 2024 during the summer were derived using the temperature inversion research technique Fig. 6. In several regions, the temperature plummeted to a minimum of −14 °C as a result of cloud cover. Certain regions see cloud cover, with the minimum recorded temperature of −14 °C. The inverted surface temperature distribution maps were further partitioned into grids using the method outlined in Section 2.2.2 The study of data disclosed the subsequent findings: In 2001, the predominant temperature range in Luoyang’s metropolitan region was between 0 °C and 30.5 °C, encompassing 46.77% of the overall urban expanse. In 2016, the predominant temperature range in the urban region was between 30.5 °C and 35.5 °C, including 53.50% of the overall urban expanse. In 2019, the predominant temperature range in the urban region was between 35.5 °C and 39.5 °C, including 51.93% of the total urban area. In 2024, the predominant temperature range in the urban region was between 39.5 °C and 43.5 °C, including 45.70% of the total urban area. In 2001, the average temperature of Luoyang’s urban area was 31.2 °C; in 2016, it was 32.01 °C; in 2019, it reached 38.07 °C; and in 2024, it increased to 40.24 °C. Between 2001 and 2024, the average temperature in Luoyang’s urban area rose by 9.04 °C.

The urban area of Luoyang is experiencing a progressive expansion of high-temperature zones. In 2024, the maximum temperature recorded in the metropolitan area was 58 °C, while the majority of the area experienced temperatures ranging from 39.5 °C to 45.5 °C. Regions with temperatures over 45 °C were predominantly located in two areas: the historical city zone north of the Luo River and the present-day Luolong District. The historic district north of the Luo River, being the original urban nucleus of Luoyang, exhibits a significant prevalence of impervious surfaces. This, along with the density of impervious surfaces from subsequent urban expansion, renders the area susceptible to excessive summer temperatures. From 2001 to 2024, Luoyang’s Luolong District had substantial expansion, characterized by extensive urban construction that resulted in a marked rise in impervious surfaces, consequently intensifying the UHI effect and elevating temperatures. By 2024, Luolong District emerged as the new urban center of Luoyang, signifying a relocation of the city’s core to the south. The Longmen Grottoes are situated in the southern region of Luolong District. The intensification of the UHI effect in southern Luoyang is anticipated to adversely affect the thermal environment of the Longmen Grottoes site, hence heightening the dangers to the preservation of the grottoes and cultural artifacts.

SUHI profile and SUHIIU-R magnitude and trend

Utilizing temperature inversion and urban-rural gradient analysis techniques, surface temperature profile charts and heat island intensity maps for Luoyang between the summers of 2001 and 2024 were generated Fig. 7. Figure 7a reveals that from 2001 to 2024, the urban surface temperature of Luoyang has markedly grown, with the most pronounced rise occurring in the Xigong District’s core, where the temperature escalated by over 12 °C. The Longmen Grottoes location underwent a temperature rise of 4 °C. The examination of Fig. 7b reveals that, over the past two decades, the intensity of the UHI effect in Luoyang has markedly intensified. The region exhibiting the most significant rise in heat island intensity was the center of Xigong District, with a maximum increase of 15 °C, succeeded by the center of Luoyang’s Luolong District, which had a 10 °C increase in heat island intensity. Between 2001 and 2024, the UHI effect caused the heat island intensity of the Jian River in the city center to rise from −2.5 °C to 2 °C, signifying the eradication of the river’s cold island effect. The heat island intensity of the Luo River rose from −4.8 °C to −2.5 °C, signifying a diminishment of the cold island effect in the Luo River. As of 2024, the heat island intensity of the Yi River near the Longmen Grottoes has exhibited no change, and the cold island effect persists. This indicates that the present UHI effect has a comparatively minor influence on the Yi River, which traverses the Longmen Grottoes site.

Fig. 7: Land surface temperature and urban heat island intensity in Luoyang city.
Fig. 7: Land surface temperature and urban heat island intensity in Luoyang city.
Full size image

a Land surface temperature of Luoyang city in 2001 and 2024. b Urban heat island intensity of Luoyang city in 2001 and 2024.

The influence of land use type changes

Figure 8 illustrates that in 2001, the principal metropolitan area of Luoyang, situated north of the Luo River, had dispersed impervious surfaces in the southern region of the river, which was predominantly comprised of villages. By 2024, advancements in Luolong District resulted in an elevated impervious surface density in the southern region of the Luo River, while the area adjacent to the southern bank of the Yi River experienced a substantial rise in impervious surfaces due to the establishment of the Yibin New District. A composite histogram was generated by doing statistical analysis on the land use classification areas throughout the four time periods. Figure 9 indicates that from 2001 to 2024, the impervious area surrounding the heritage property expanded from 139.62 km² to 399.30 km2, and vegetation covering significantly declined from 639.88 km2 to 390.35 km2. Between 2001 and 2024, the vegetation covering rate in the vicinity of the site diminished from 81.06% to 48.24%, while the fraction of impervious surfaces escalated from 17.48% to 49.45%.

Fig. 8: Land use type classification maps and satellite images of Luoyang city.
Fig. 8: Land use type classification maps and satellite images of Luoyang city.
Full size image

a–c Land usetype classification maps of the urban area of Luoyang city in 2001, 2016, 2019, and 2024, respectively. e–h Satellite images of the urban area of Luoyang city in 2001, 2016, 2019, and 2024, respectively.

Fig. 9
Fig. 9
Full size image

Land use type classification maps of Luoyang city for the years 2001, 2016, 2019, and 2024.

The northeastern section of the Longmen Grottoes, now transformed into a parking lot, and the western mountain housing significant attractions like the Lushena Buddha were designated as the research area. To comprehensively assess the influence of land use alterations on heat island intensity at the Longmen Grottoes site, the temperature inversion technique and urban-rural gradient research method were employed to derive the average heat island intensity for the western mountainous region of the Longmen Grottoes site during the years 2004, 2008, 2014, 2016, 2020, 2022, and 2023 Fig. 10. The seven-year dataset on heat island intensity for the western mountain of the Longmen Grottoes site is as follows: −4.67 °C, −5.06 °C, −1.24 °C, −1.46 °C, 0.67 °C, 1.31 °C, 2.46 °C. Simultaneously, the 30-meter resolution annual land cover data from 1985 to 2023, provided by Professors Yang Jie and Huang Xin of Wuhan University, was utilized to acquire the impervious surface density and green space density data for the metropolitan area of Luoyang over the seven-year period. The impervious surface density for the urban area of Luoyang during seven years were as follows: 24.91%, 29.15%, 36.66%, 38.02%, 39.87%, 40.92%, and 41.11%. The green space surface density for the metropolitan area of Luoyang for the seven years was: 73.77%, 69.39%, 61.43%, 59.96%, 58.17%, 57.26%, and 57.11%. A subsequent study was undertaken to examine the correlation between heat island intensity in the heritage site and variations in impervious surface and green space density in the urban area of Luoyang. A linear regression analysis was conducted utilizing heat island intensity, impervious surface density, and green space density across these seven years Fig. 11.

Fig. 10: Satellite image and land surface temperature map of the Longmen Grottoes area.
Fig. 10: Satellite image and land surface temperature map of the Longmen Grottoes area.
Full size image

a Satellite map of the Longmen Grottoes. b–h show the land surface temperature maps of the Longmen Grottoes area for the years 2001, 2004, 2008, 2016, 2020, 2022, and 2023, respectively.

Fig. 11: UHI intensity in the Longmen Grottoes site area and impervious surface and green space maps of Luoyang city.
Fig. 11: UHI intensity in the Longmen Grottoes site area and impervious surface and green space maps of Luoyang city.
Full size image

a Variation of UHI intensity in the Longmen Grottoes area. b Relationship between UHI intensity in the Longmen Grottoes area and impervious surface density. c Relationship between UHI intensity in the Longmen Grottoes area and green space density.

The findings demonstrate a substantial positive association between the average heat island intensity and the density of impervious surfaces in the Longmen Grottoes site area. A notable negative association exists between the average heat island intensity and green space density in the Longmen Grottoes site area (p < 0.01). Figure 11 illustrates that alterations in green space density similarly affect the UHI intensity at the Longmen Grottoes site as do changes in impervious surface density. The slope values for green space density and impervious surface density are −41.15 and 42.89, respectively. This finding indicates that the UHI effect in Luoyang has adversely impacted the thermal environment of the Longmen Grottoes site area. Consequently, the influence of UHI effect intensity on adjacent heritage monuments need careful consideration. An effective urban land use planning approach is crucial for the preservation of cultural sites in urban regions.

Discussion

This research builds upon extensive prior studies to investigate the influence of UHI effects on heritage sites situated on the urban periphery. Alterations in the thermal environment of heritage site locations are frequently affected by fluctuations in the local microenvironment and climatic conditions. Previous researches have indicated that variations in humidity and temperature within caves generally influence the modification of soluble salts or result in intricate salt damage to the cave structures22. Consequently, comprehending the effects of local climate alterations on the preservation of heritage sites is essential. This is particularly applicable to heritage sites on the urban periphery, as their surroundings are more vulnerable to the UHI effect resulting from urban expansion. This study revealed that the UHI effect has progressively strengthened due to urban expansion, resulting in an overall increase in heat island intensity that impacts the vicinity of heritage sites. This illustrates the continuous effect of swift urban development on the environment of cultural sites located on the urban periphery. The cumulative effect of recurrent environmental disturbances can result in alterations to cultural heritage, frequently yielding associated and irreversible outcomes10,11,12. Previous studies have neglected the influence of the UHI effect on the thermal conditions of heritage sites located on the urban periphery, resulting in inadequate protective measures for these areas.

Bi et al. 22 created a one-dimensional model for the coupling of heat and moisture transmission, utilizing temperature and vapor pressure as driving forces. Experimental and numerical assessments were performed on the moisture and heat transmission within the cave walls of the Mogao Grottoes. The Getty Conservation Institute refined and authenticated the photogrammetric condition monitoring procedure, enhancing existing mural monitoring methodologies52. Certain scholars have executed meticulous surveillance of the microenvironment in heritage site regions to safeguard the murals. This monitoring encompasses carbon dioxide concentration, particulate matter levels, and ventilation53. Nevertheless, the aforementioned studies predominantly concentrate on the microenvironment of heritage sites and the methodologies for monitoring site conditions, neglecting to incorporate study from the standpoint of local climatic factors. In summary, alterations in the local climatic environment are a pivotal component in examining the transformations of urban heritage site contexts. This study seeks to investigate the influence of the UHI phenomenon on the thermal conditions of heritage sites located on the outskirts of urban areas. This study examines land surface temperature and UHI intensity over several eras by acquiring Landsat data and employing temperature inversion techniques. The study analyzes the variations in the severity of the heat island effect from the urban center to the adjacent heritage site region, based on urban profile assessment. This framework offers a more thorough and dependable local environmental assessment for the proactive safeguarding of heritage sites on the urban periphery.

This study examined the influence of the UHI effect on the thermal environment of heritage sites in urban areas and conducted a robust analysis, establishing a significant linear relationship between changes in land use type proportions in urban areas and UHI intensity surrounding heritage sites. The ratio of green space in urban environments is inversely associated to UHI intensity near heritage sites, but the ratio of impervious surfaces is directly connected to UHI intensity. Future urban planning in Luoyang should focus enhancing urban greening rates based on these observations. The vicinity of the Longmen Grottoes should minimize the installation of impermeable surfaces and refrain from industrial edifices that hinder heat dispersion. In the heritage site, suitable landscape alterations in the southern section of the Longmen Grottoes could enhance vegetation coverage.

Moreover, the results of this study offer a novel perspective for the preservation of heritage assets in various cities globally. Other cities with nearby heritage sites should evaluate the impact of UHIs on their surrounding landscapes. It might be beneficial to examine whether the pyramids in Cairo, Egypt, or the Yungang Grottoes near Datong, China, are affected by variations in UHI intensity. The pyramid site in Egypt is situated about 13 kilometers from Cairo’s city center, and the impact of urban expansion and human activities in Cairo during recent decades on the pyramid environment warrants more investigation. Future research ought to concentrate on the integration of terrestrial measurements with remote sensing techniques, alongside enhancing temporal resolution for synchronized monitoring of the heritage site and its adjacent surroundings. Furthermore, it is essential to implement the research methodology in additional urban heritage site regions, taking into account more extensive contextual factors to investigate the diverse effects of UHI on these locations. Enhancing the assessment of UHI intensity on adjacent cultural sites will yield more comprehensive and dependable local environmental analyses, facilitating preventive conservation measures for these sites and fostering sustainable urban development that harmonizes the preservation of human civilization.

This study contains aspects that could be enhanced in subsequent research. Due to adverse weather circumstances, appropriate satellite remote sensing data for July 2024 was inaccessible; instead, the study utilized data from June of that year, which may have resulted in an underestimating of UHI intensity, thereby bringing potential inaccuracies into the analysis. The study primarily examined alterations in surface temperature during the past 20 years during the summer, the season exhibiting the most significant UHI intensity, and assessed the current winter surface temperatures. Nevertheless, spring and fall were excluded from the study, and subsequent research could examine these seasonal differences more comprehensively. The absence of data for other seasons, including winter, spring, and autumn, constrains the thoroughness of the findings. The unique characteristics of the cultural site have complicated the balance between preservation and tourism, hindering the execution of numerous experiments. Nonetheless, we have conducted a substantial number of experiments. Long-term, multi-point field monitoring has yet to be achieved, and it is essential to undertake such monitoring in the future to acquire a more profound and multidimensional comprehension of the environmental conditions within the cultural site area.