Climatic effects on landscape multifunctionality in urban parks: a view for integrating ecological supply and human benefits

The rapid and relentless development of urban areas highlights the importance of landscape multifunctionality. However, there is limited research on the temporal dynamics and climatic effects of urban landscape multifunctionality. This study aimed to address this gap by analyzing the features of multiple landscape functions triggered by seasonal climate change in different urban park types. In this study, we investigated five typical urban landscape functions (alleviating urban heat islands, vegetation growth, biodiversity promotion, alleviation of waterlogging, and provision of recreational activities) by establishing a set of indices: ecological supply capability (SP ), proportion of ecological supply (SPP ), capability of human benefits (BP ), and human benefits efficiency (BEP ) of urban parks. The average SP of the landscape functions was 58% in summer and 46% in winter. During the transition from summer to winter, urban parks witnessed a significant decrease in SPP for alleviating the urban heat island, dropping from 34% to 5%. The primary landscape functions shifted from alleviating the urban heat island (34%) and providing recreation (29%) to providing recreation (38%) and biodiversity promotion (29%). Concerning park types, nature parks provided the highest SP , whereas community parks provided the highest BEP . This study has useful implications for landscape management in urban parks, particularly regarding timely adjustments across seasonal climates. It is possible to promote sustainable and effective human well-being by maximizing landscape functions.


Introduction
Rapid urbanization is associated with ecological issues that are closely linked to the well-being of urban residents (Yang et al 2023).In highly developed urban areas, it is crucial for current developmental strategy to shift the focus from solely considering land use and development potential to evaluating the capacity of landscape functions (Scott et al 2016).Landscapes represent the interaction between the primary structure, land use and other human activity, and landscape functions are guided by various drivers with high temporal and spatial heterogeneity (Bastian et al 2006, Bolliger et al 2011).Landscape multifunctionality facilitates intricate interactions between multiple land units and stakeholders, leading to a wide array of societal benefits, which are usually conceived and evaluated as a joint supply of various ecosystem services at the landscape level (Liang et al 2020).Understanding the landscape multifunctionality can help in developing the landscapes that span across urban and rural areas, planning these landscapes in an integrated manner, and providing adaptive climate change strategies, thereby making land use more sustainable (Selman 2009, Galler et al 2015).
With growing urban populations and increasing land pressure, urban landscape multifunctionality has received increasing attention in the recent years (Belmeziti et al 2018, Charoenkit andPiyathamrongchai 2019).The choice of landscape functions for assessing landscape multifunctionality varies based on the study area characteristics, and hence, there are several types of landscape functions, assessment methods, and scoring systems (Charoenkit and Piyathamrongchai 2019).For instance, mountainous areas have five typical landscape functions: net primary productivity, soil retention, water yield, crop production, and residential support (Peng et al 2019).Agricultural areas have landscape aesthetics function (Peng et al 2017), whereas mining landscapes have soil conservation function (Zhang et al 2022).Landscape planning can be significantly enhanced if landscape multifunctionality prioritizes the well-being of people (Mastrangelo et al 2014).In highly urbanized areas, the various methods for assessing landscape functions should consider the daily life of residents.Examples include drainage management, habitat provision, ecological connectivity, health and well-being, recreational space, energy reduction, and mitigation and adaptation of climate change (Scott et al 2016).
Current research on landscape multifunctionality primarily focuses on large scales, and assessments of landscape multifunctionality in urban areas with high population concentration are lacking.Few studies have focused on the temporal-scale changes in landscape multifunctionality and annual changes (but these studies have low temporal precision) (Firozjaei et al 2020, Li et al 2020, Zhang et al 2020b).Similarly, studies on temporal changes in temperature, population distribution, and the supply and benefits of landscape functions are limited.Hence, it is essential to enhance our understanding of the alterations in landscape functions due to seasonal as well as the subsequent impacts on ecological supply and human benefits.
The promotion of landscape multifunctionality has become a common goal of land use policies (Mastrangelo et al 2014, Liang et al 2020).In urban landscapes, parks serve multiple functions by providing passive and active recreation options, environmental benefits, and wildlife habitats (Solecki and Welch 1995).However, the space allocated to urban parks in high-density cities is limited; hence, smaller parks must perform more landscape functions to meet the needs of urban residents.This necessitates that changes in landscape functions of different types of urban parks must be studied under seasonal climatic changes.This can aid in the application of dynamic landscape management models to promote a more sustainable, efficient, and integrated landscape.This requires a precise understanding of the ecological supply and human benefit potential of urban parks at different times to provide a basis for a dynamic management model.
In cities, ecological problems that affect urban residents include urban heat islands and waterlogging (Awal 2014, Tian et al 2021, Zhao and Yue 2023).An impact of urban heat islands on human health is significant, particularly when considering the potential impacts of climate change (Ward et al 2016, Heaviside et al 2017), whereas waterlogging can cause considerable human casualties, economic losses, and social issues (Yin et al 2015).In addition, the conservation of biodiversity in urban environments is becoming increasingly urgent, and urban landscapes are important for regional and global biodiversity (Elander et al 2005, Dearborn andKark 2010).Vegetation is an important part of the urban space, and its growth significantly affects the quality of the urban landscape (Smardon 1988, You andPan 2020).Recreation is one of the basic needs of urban residents, and its accessibility is one of its most important factors (Rigolon 2016, Xu et al 2017).These five landscape functions represent the core requirements of urban residents.
The study focused on urban parks as a subject and identified five typical landscape functions: urban heat island alleviation, vegetation growth, biodiversity promotion, waterlogging reduction and providing recreation activities.The two main objectives of the present study included: (1) assessing the landscape functions of urban parks and (2) examining the temporal variation characteristics of landscape multifunctionality in the ecological supply and human benefits of urban parks.

Study area and urban park types
Beijing is the capital of China and has a typical temperate monsoon climate, with a hot, rainy summer (the hottest month is July) and a cold, windy winter (the coldest month is January) (An et al 2021, Zhang et al 2021a).Beijing has experienced rapid urbanization since the 21st century (Li et al 2012, Liu et al 2020).In the present study, we included the central urban areas as defined in the 'Beijing Urban Master Plan (2016-2035)' .These areas include the districts of Dongcheng, Xicheng, Chaoyang, Haidian, Fengtai, and Shijingshan.These areas account for more than half of the total population of Beijing, and represent the most important urban areas, including national political center, cultural center, international exchange center, and scientific and technological innovation center in China (Zhang et al 2012, 2018, 2021b, Beijing Planning and Land Resources Management Committee 2017).
Urban parks were the primary study subjects for landscape multifunctionality, whereas non-park urban areas were selected for comparison (figure 1).The total study area is 1384 km 2 , of which 136 km 2 is covered by urban parks, representing approximately 10% of the total study area, and 1248 km 2 is nonpark urban areas, representing approximately 90% of the study area.The directory of the parks in Beijing was obtained from the official website of the Beijing Municipal Bureau of Landscape and Greenery (http:// yllhj.beijing.gov.cn/English/Resources/).A total of 537 parks were mapped in ArcGIS 10.8 based on park boundaries obtained from the Baidu map (https:// map.baidu.com).The types of parks were defined based on the 'Beijing Municipal Parks Categorization and Grading Management Measures' issued in 2022 (table 1).

The assessment framework of ecological supply and human benefits
In this study, we focused on the landscape functions of urban parks within five typical aspects: alleviation of urban heat islands, vegetation growth, biodiversity promotion, alleviation of waterlogging, and provision of recreational activities.Specifically, ecological supply and human benefits were assessed using the following steps (figure 2).Here, S P is the ecological supply capability of urban parks; SP P is the proportion of ecological supply capability of urban parks; B P denotes the human benefits of urban parks; BE P denotes the human benefit efficiency of urban parks.

Ecological supply capability (S P )
The ecological supply capability of landscape multifunctionality was calculated using the ratio of the landscape function actual improvement, the maximum potential improvement of urban parks compared to non-park urban areas or the entire study area, and a value (S UHI , S VG , S B , S W , S R ) between 0 and 1.Among them, we considered a full score of S UHI when the regulating capacity of vegetation was ⩾3.2 • C in summer and ⩾1 • C in winter, which was based on a previous long-term study (Yan et al 2023a).The variables for each landscape function were calculated using the following equations: LST OP − LST P 3.2 (summer) or 1 (winter) (1) Here, S P is the ecological supply capability of urban parks, that is the proportion that reaches the   The details of the method and data sources for land surface temperature (LST), normalized difference red edge index (NDRE), species richness index (D), runoff volume (Q), and cover ratio (CR) can be found in section 2.3.

Proportion ecological supply capability (SP P )
where, SP P is the proportion of ecological supply capability of urban parks.

Human benefits (B P )
where, B P denotes the human benefits of urban parks, and P P denotes the population in urban parks, which is obtained based on the mobile phone signaling data of the top three operators in China (China Mobile, China Unicom, and China Telecom).

Human benefits efficiency (BE P )
BE P = B p A p (9) where BE P denotes the human benefit efficiency of urban parks, A P denotes the area of urban parks (km 2 ), and D P denotes the population density of urban parks (people km −2 ).

Data source and methods for assessing multiple landscape functions 2.3.1. Urban heat island
The urban heat island was characterized using LST based on the Google Earth Engine (GEE) and inversion LST using the method described by Ermida et al (2020).This method obtained data from Landsat 4/5/7/8 satellite and used the Statistical Mono-Window algorithm developed and used by the Climate Monitoring Satellite Application Facility to derive LST data records from the Meteosat First-and Second-Generation series of satellites.The synthetic values were calculated for 2021 using the following equation: where, Tb is the top-of-atmosphere brightness temperature in the thermal infrared channel of LANDSAT, and ε is the surface emissivity for the same channel.The algorithm coefficients A i , B i ; and C i were determined from linear regressions of radiative transfer simulations performed for ten classes of total column water vapor (TCWV), ranging from 0 to 6 cm with each having a difference of 0.6 cm and values of TCWV > 6 cm being assigned to the last class.

Vegetation growth
where, D is the species richness index, A is the unit area, and S is the total number of species in the community.where I a is the initial rainfall loss (mm); there is an empirical relationship between I a and S. Given the urban characteristics of the study area, a value of λ = 0.05 is employed in this context (Woodward et al 2003).

Waterlogging
where P is the average monthly summer/winter rainfall total (mm) and Q is the runoff volume (mm).Daily rainfall data were obtained from the Resource and Environment Science Data Center at the Chinese Academy of Sciences.

Recreational activities
The recreational activity function is determined by the gravity index, which was first developed by Hansen (1959).The field strength of an urban park is calculated as follows: where F k is the field strength of the kth park; a k is the area of the kth park (m 2 ); and d k is the gravitational impact distance from the kth park (m).The study was based on 692 questionnaires in Beijing in 2022 (Yan et al 2023b), and collected information on the maximum acceptable distance between pockets, communities, and other parks (Yan, unpub.data) (table 3).The field strength was found to be 0.03 for pocket and community parks and 0.15 for other parks.The gravity impact distance of different parks was calculated based on park area and field strength and used as an indicator of the recreational activity function of individual parks.The CR of the buffer area to the attraction distance radius was used as an indicator of park type.

Landscape functions of urban parks and non-park urban areas
The landscape functions of urban parks were significantly higher than those of non-park urban areas (figure 3).In summer, the landscape functions of alleviating urban heat island, improving vegetation growth, and alleviating waterlogging are more significant, whereas in winter, biodiversity promotion is more significant.The average LST of urban parks was 3.7 • C lower than of non-park urban areas in summer while this difference reduced to 1.5 • C in winter, means that the parks have cooler temperatures, especially in the summer.The NDRE of urban parks was 0.2 higher in summer and 0.1 higher in winter than that of non-park urban areas, indicating a denser vegetation cover and better growth in parks than in non-park urban areas.The species richness index was 0.7 higher in summer and 1.1 higher in winter in non-park urban areas and showed richer biodiversity in the parks.In terms of urban waterlogging, runoff volume in the parks was 28.1 mm lower than non-park urban areas in the summer and 2.0 mm lower in the winter, suggesting that waterlogging is less likely to occur in the park during the rainy season.
The recreation function shows the overall coverage of urban parks is currently 85%.Except for recreational functions, the landscape functions of urban parks were significantly different between summer and winter (figure 4).During the same season, disparities can be also observed among various urban parks, particularly in terms of biodiversity promotion and alleviation of waterlogging, which exhibit relatively greater variation.In summer, historical parks are particularly effective in alleviating the urban heat island effect, while nature parks excel in enhancing vegetation growth, promoting biodiversity, and alleviating waterlogging.Community parks, on the other hand, are the best for providing recreational activities.In winter, nature parks emerge as the optimal choice for vegetation growth, biodiversity promotion, and alleviating waterlogging, while community parks are still the best for providing recreational activities.

Seasonal changes in ecological supply
The amount of ecological supply extent (S P ) in summer and winter of all parks was 2.9 and 2.3, respectively (table 4).This demonstrates that the average S P of the five landscape functions was 58% and 46% in summer and winter, respectively.This represents the extent to which urban parks attain their maximum ecological supply potential.Nature parks had the highest S P in both summer and winter, whereas pocket parks had the lowest S P .The S P of comprehensive parks decreased the most from summer to winter, whereas those of natural parks decreased the least.
The results show that the main proportion of ecological supply capability (SP P ) changed from alleviating urban heat island (SP P :34%) and recreation (SP P :29%) in summer to recreation (SP P :38%) and biodiversity promotion (SP P :29%) in winter (figure 5(a)).There was a sharp decrease from summer to winter (from 34% to 5%) in the SP P of urban parks for alleviate the urban heat island effect.There was also no significant change in the SP P for vegetation growth, and there was an improvement in the alleviation of waterlogging.The results of SP P changes of different types of urban parks show that the main supply capability of comprehensive, theme, ecological, and nature parks changed from alleviating the urban heat island in summer (SP P :51%, 42%, 49%, 43%) to alleviating waterlogging (SP P :58%, 64%, 61%, and 50%), and community parks have changed from alleviating the urban heat island in summer (SP P :35%) to recreation in winter (SP P :42%).The SP P of historical and pocket parks were the same in summer and winter, and the primary supply capability of historical parks was to alleviate the urban heat island (summer SP P :63%; winter SP P :47%) while for pocket parks recreation was the primary supply capability (summer SP P :35%; winter SP P :57%) (figure 5(b)).

Temporal variation in human benefits
We determined the number of visitors to each type of park (P P ) on working and rest days in summer and winter (figure 6).The results showed that the P P in each park type was relatively high on winter rest days and lowest mostly on summer rest days.
The highest number of people in theme parks, and the lowest in nature parks.The human benefits (B P ) were higher in theme parks and lower in nature parks.
The human benefits efficiency (BE P ) for community parks was relatively high, especially during weekdays in summer, followed by theme, pocket, and comprehensive parks.

Potential mechanism of urban landscape multifunctionality
This study integrates various analytical methods, such as remote sensing inversion, and model simulation, combined with cell phone signaling data, to quantitatively portray the characteristics of the ecological supply and human benefits of urban parks with seasonal climate change and further reveal the role of urban parks in fulfilling landscape functions.
The results demonstrated that the landscape functions of urban parks were enhanced compared to those of non-park urban areas.In summer, it is more significant to alleviate urban heat island, enhance vegetation growth, and mitigate urban waterlogging, whereas biodiversity promotion is more significant in winter.The S P of nature parks was more significant than that of other types of parks; the S P of pocket parks was relatively low, except for the increase in recreational functions, indicating that it is not feasible to increase ecological benefits through the construction of pocket parks.Related studies have indicated that comprehensive parks have the best cooling effect, and that small urban parks may be relatively fragile in terms of biodiversity promotion (Amaya-Espinel et al 2019, Peng et al 2021).The study discovered that historical and nature parks exhibited higher biodiversity, potentially due to the diverse vegetation composition found in historical parks, and the distance from nature parks to city center (Yang et al 2020).Nature parks have the best cooling effect, which may be related to vegetation cover and size (Cheng et al 2015, Yang et al 2017).Furthermore, we found that the collective impact of parks on biodiversity conservation and recreation was far more significant than that of individual parks.This highlights the need to prioritize the synergy of park conservation during park development.
We observed that the SP P of the urban parks changed significantly from summer to winter.The main landscape functions were to alleviate the urban heat island effect during summer, whereas pocket parks prioritized recreational functions.This is likely due to the smaller size of the pocket parks.Furthermore, pocket parks tend to serve recreational functions, primarily for nearby residents.In winter, most parks have lost their heat island mitigation function because of the background temperature, and the dominant function of parks has shifted.This may be related to the location, or it may be because hot weather reduces the willingness of urban residents to travel.Community and pocket parks had the highest population densities; however, the BE P of community parks was higher than that of pocket parks.This shows that, for high-density built-up areas, community parks are more in providing more convenient landscape functions to urban residents.Physical exercise and recreation are the most common reasons for using parks; therefore, some studies suggest providing parks within 1000 m of residential areas, which illustrates the importance of community parks (Liu et al 2017).

Implications for urban landscape management
First, an assessment of landscape multifunctionality offers a more comprehensive understanding of the role of urban parks.This study extends beyond the comparison between urban and non-park urban areas.Moreover, it is crucial to consider supply capabilities across different periods when assessing the provision of landscape functions in parks.Landscape functions undergo temporal variations.Therefore, in the evaluation of urban ecosystems and landscape management, it is essential to establish distinct index monitoring systems for different periods to enhance accuracy.In addition, it is possible to organize activities that align with seasonal changes in landscape functions.It allows urban residents more deeply appreciate the ecological value of urban parks.Based on the research findings, urban residents can be encouraged to visit parks for recreation during summer, whereas in winter, activities focused on animal protection and bird watching should be emphasized.In essence, the assessment of urban landscape multifunctionality contributes to the holistic and sustainable development of cities, fostering harmonious coexistence between human activities and the natural environment, while promoting the well-being of urban populations.
Assessing urban landscape multifunctionality has significant implications for enhancing climate change resilience in urban environments.By evaluating the diverse functions of urban parks, particularly their ecological and adaptive capacities, cities can prepare for and mitigate the impacts of climate change.Overall, this study provides a roadmap for cities to enhance their climate change resilience by strategically incorporating natural and green infrastructure elements into urban planning and design.This holistic approach contributes to the long-term sustainability and adaptability of urban areas that face changing climates.
Second, we recognize that the dynamics of S P are strongly influenced by the background climate.Current global warming trends may lead to an increase in urban heat islands (Sachindra et al 2016), an increase in the severity of waterlogging events (Zhang et al 2020a, Ding et al 2022), a decrease in biodiversity (Nunez et al 2019), and the promotion of vegetation growth (Ren et al 2022).These variations in landscape functions can lead to changes in SP P , which in turn can affect the benefits that urban residents derive from urban parks.This means that we need to make changes to the construction and management of parks according to changes in the climate to continuously maintain and improve the health and well-being of urban residents.The types and layouts of parks can be adjusted according to the efficiency of their ecological supply and human benefits.For example, community parks should be strengthened.The impact of climate change on the landscape multifunctionality of urban parks must be explored further to make timely adjustments to landscape management strategies.
Lastly, considering the number of visitors to each park type and human benefits, we can establish a more effective management plan for urban park types.The main approach of the current urban greening program is to prioritize the land vacated by functional changes, as the high-density development of cities makes urban ecological restoration very difficult.In contrast, the efficiency of urban park services can be improved by transforming park types based on the results.We found that community parks have the largest benefit efficiency, implying that they occupy an important share of the urban park system, and should focus on strengthening the construction intensity and subsequent maintenance of community parks.

Limitations and future improvements
Despite conducting a comprehensive analysis of the landscape multifunctionality of urban parks, this study acknowledges its limitations.First, the present landscape multifunctionality assessment focused on the central area of the city.Thus, the selection of landscape multifunctionality considers indicators that are closely related to urban residents.In terms of quantifying the ecological supply capacity, non-park urban areas are mainly used for comparison, which may have certain regional limitations.Second, in terms of data collection, although remote sensing can obtain LST data on a large scale, there is still a certain difference compared to the actual temperature perceived by urban residents.There is also a need to expand the scope of biodiversity monitoring to ensure comprehensive coverage, as the species richness of small parks, such as pockets and community parks, may be underestimated due to limitations in data collection.In addition, when assessing recreational functions in parks, emphasis is placed on the accessibility of urban parks, overlooking the influence of infrastructure (such as sports fields and fitness equipment).Future research should quantify the purpose of park visits by urban residents using questionnaires and other methods.This could enhance the degree of matching between landscape functions and residents' benefit demands.Furthermore, big data can be used to build comprehensive urban park evaluation systems to better understand landscape function values.In addition, we can strengthen further research on the range of influence of landscape multifunctionality in urban parks and non-park urban areas.

Conclusions
This study investigated landscape multifunctionality in both urban and non-urban parks of Beijing, a representative Chinese metropolis.This study analyzed the variations in the supply and benefits of landscape functions in urban parks using remote sensing inversion and ecological model.It quantified the ecological supply and human benefit changes in landscape functions attributed to population dynamics on working and rest days during summer and winter.From the point of view of S P , nature parks have high ecological value and are the best places for urban residents, while from the point of view of BE P , it is the most cost-effective for the government to increase the construction of community parks.The study results can better understand the supply and benefits of landscape multifunctionality and its temporal changes and explore the coupling characteristics of people and landscape functions of different urban park types.These findings are instrumental in maximizing the landscape multifunctionality of urban parks, improving their allocation of urban parks and promoting sustainable human well-being.
gardens with outstanding historical and cultural values that have had an impact on urban change or the development of culture and art.Most of the trees in parks are old and tree coverage is higher.Comprehensive 44 Comprehensive parks are parks that are well-functioning and able to meet the needs of a wide range of activities such as excursions, recreation, popular science and culture.Most of them are newly built parks in recent decades, generally with open lawns and special fitness trails.The minimum area ⩾ 5 hm 2 .Community 148 Community parks are mainly used for daily recreational activities by the residents in the neighborhood.There are necessary supporting facilities and activity venues, and the minimum area ⩾ 0.5 hm 2 .Low tree cover in most parks.Theme 80 Theme parks refer to a park distinctive thematic content, such as botanical gardens, zoos.Pocket 222 Pocket parks are located within the scope of urban construction land, land independent, small-scale parks, and most of the trees are short in age.Ecological 21 Ecological parks are located outside the scope of urban construction land, considering the multiple functions of public recreation, ecological environmental protection, such as country parks.High tree cover in most parks.Nature 3 Nature parks refer to areas in the system of nature reserves that are open to the public and have the functions of recreation as well as popularization of education, such as forest parks, geoparks, wetland parks, and so on.Most of the trees in parks are very tall and has a very high tree cover.

Figure 2 .
Figure 2. The assessment framework for characterizing temporal changes in the ecological supply and human benefits.

Figure 3 .
Figure 3. Landscape multifunctionality assessment of urban parks.(a) Urban heat island, characterized by LST; (b) vegetation growth, characterized by NDRE; (c) biodiversity, characterized by species richness index; (d) waterlogging, characterized by runoff volume; and (e) recreation activities, characterized by cover ratio of urban parks.

Figure 4 .
Figure 4. Landscape multifunctionality assessment of different types of urban parks (the error line indicates the standard deviation, and the gray and yellow dashed lines indicate the landscape functions of all urban parks in summer and winter, respectively).

M Yan et alFigure 6 .
Figure 6.PP, DP, BP and BEP of different park types (total amount of a particular park type was indicated).

Table 1 .
Urban parks type statistics.

Table 2 .
Curve number (CN) values corresponding to each land cover type (AMI I).

Table 3 .
Questionnaires of maximum acceptable distance.

Table 4 .
The SP for different urban park types in summer and winter.
Figure 5.The SPP of different urban park types in summer and winter.