Asynchronous modes of paleofire occurrence in eastern monsoonal China since the LGM ^("え "){ }^{\text {え }} 末次盛冰期以来中国东部季风区古火灾发生的异步模式 ^("え "){ }^{\text {え }}
Menglin Song ^(a,b,c){ }^{\mathrm{a}, \mathrm{b}, \mathrm{c}}, John Dodson ^(a,b,^(**)){ }^{\mathrm{a}, \mathrm{b},{ }^{*}}, Scott Mooney ^(b){ }^{\mathrm{b}}, Ge Shi ^(a,d){ }^{\mathrm{a}, \mathrm{d}}, Hong Yan ^(a,e){ }^{\mathrm{a}, \mathrm{e}} 宋梦琳 ^(a,b,c){ }^{\mathrm{a}, \mathrm{b}, \mathrm{c}} ，约翰·多德森 ^(a,b,^(**)){ }^{\mathrm{a}, \mathrm{b},{ }^{*}} ，斯科特·穆尼 ^(b){ }^{\mathrm{b}} ，石戈 ^(a,d){ }^{\mathrm{a}, \mathrm{d}} ，严宏 ^(a,e){ }^{\mathrm{a}, \mathrm{e}}^("a "){ }^{\text {a }} State Key Laboratory of Loess Science, Institute of Earth Environment, Chinese Academy of Sciences, Xi'an, Shaanxi 710061, China ^("a "){ }^{\text {a }} 中国科学院地球环境研究所 黄土科学国家重点实验室，陕西 西安 710061^(b){ }^{\mathrm{b}} School of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney 2052, Australia ^(b){ }^{\mathrm{b}} 新南威尔士大学生物、地球与环境科学学院，悉尼 2052，澳大利亚^(c){ }^{\mathrm{c}} University of Chinese Academy of Sciences, Beijing 100049, China ^(c){ }^{\mathrm{c}} 中国科学院大学，北京 100049，中国^(d){ }^{\mathrm{d}} Xi'an Institute for Innovative Earth Environment Research, Xi'an 710061, China ^(d){ }^{\mathrm{d}} 西安创新地球环境研究所，陕西 西安 710061^(e){ }^{\mathrm{e}} Institute of Global Environmental Change, Xi'an Jiaotong University, Xi'an 710049, China ^(e){ }^{\mathrm{e}} 西安交通大学全球环境变化研究院，西安 710049，中国
A R T I C L E I N F O 文章信息
Editor: Dr. Jed O Kaplan 编辑：杰德·O·卡普兰博士
Keywords: 关键词：
Fire history 火灾历史
China monsoonal region 中国季风区
Anti-phase mode 反相模式
ENSO 厄尔尼诺-南方涛动
Abstract 摘要
Wildfire is a key environmental influence on ecosystems, global and regional climates, and human societies. Reconstructing past fire histories and investigating the relationships between wildfires, vegetation, and climate are crucial for understanding the long-term patterns of fire regimes and predicting future wildfire risks. Although there has been considerable research on local fire events, consideration of regional-scale fire over extended timescales is not well covered in China. In this study, we synthesized 70 proxy records of past fire, including charcoal and black carbon, from eastern monsoonal China, since the Last Glacial Maximum (LGM, 21-18 ka BP) to investigate the spatiotemporal patterns in fire history. Our results reveal an asynchronous pattern in fire activity between southern and northern China. In southern China, higher fire activity occurred during the LGM, Younger Dryas ( 12.8-11.6kaBP12.8-11.6 \mathrm{ka} \mathrm{BP} ) event and early Holocene ( 11-7kaBP11-7 \mathrm{ka} \mathrm{BP} ), corresponding to periods of drier climate, and suggesting that fire regimes were enhanced by lower precipitation there. In contrast, fire activity in northern China increased during the Bølling-Allerød (B/A, 14.8-12.9 ka BP) warming and the mid-Holocene, with peak activity associated with wetter periods that probably promoted plant growth, highlighting the role of biomass as fuels for fire events. Regional fires might be attributed to precipitation control modulated by regulating the combustibility and availability of fuels. Furthermore, El Niño-Southern Oscillation (ENSO) variability creates episodes of drought and rain by increasing or decreasing the intensity of the East Asian Summer Monsoon (EASM), and is therefore likely linked to wildfire dynamics. 野火是影响生态系统、全球及区域气候以及人类社会的重要环境因素。重建历史火灾记录并研究野火、植被与气候之间的相互关系，对于理解火灾模式长期演变规律和预测未来野火风险至关重要。尽管针对局部火灾事件已有大量研究，但中国在长时间尺度上对区域规模火灾的考察仍显不足。本研究综合了东亚季风区自末次盛冰期（LGM，21-18 ka BP）以来的 70 个古火灾代用指标（包括木炭和黑碳记录），以探究火灾历史的时空分布特征。研究结果揭示了中国南方与北方火灾活动存在异步变化模式。 中国南方地区的火灾活动在末次盛冰期、新仙女木事件（ 12.8-11.6kaBP12.8-11.6 \mathrm{ka} \mathrm{BP} ）和早全新世（ 11-7kaBP11-7 \mathrm{ka} \mathrm{BP} ）更为频繁，这些时期对应着较为干旱的气候条件，表明降水减少可能加剧了该地区的火灾态势。与之形成鲜明对比的是，中国北方在波令-阿勒罗德暖期（B/A，14.8-12.9 ka BP）和全新世中期的火灾活动有所增加，其峰值出现在气候湿润期——这种湿润环境可能促进了植物生长，凸显出生物量作为火灾燃料的重要作用。区域火灾活动可能受降水调控影响，降水通过改变燃料的可燃性与可获得性发挥作用。此外，厄尔尼诺-南方涛动（ENSO）通过增强或减弱东亚夏季风强度来引发干旱与多雨期，因此很可能与野火动态存在关联。
1. Introduction 1. 引言
Wildfires exert a profound impact on the environment and human societies, influencing the distribution and functioning of biomes, altering vegetation community structures and composition, reshaping the landscape, causing soil erosion and changing atmosphere composition, including gases and aerosols (Flannigan et al., 2009; Whitlock et al., 2010; Zheng et al., 2021). According to the IPCC, global climate change is expected to elevate the risk of extreme fire events, with a projected net increase in wildfire frequency of approximately 30%30 \% (IPCC, 2021). Over recent decades, severe fire events have caused widespread forest destruction, loss of life and impacted economies in regions such as Canada, Russia, the Amazon, and southern Australia 野火对环境和人类社会产生深远影响，会改变生物群系的分布与功能，破坏植被群落结构与组成，重塑地表景观，造成土壤侵蚀并改变大气成分（包括气体与气溶胶）（Flannigan 等，2009；Whitlock 等，2010；Zheng 等，2021）。根据政府间气候变化专门委员会（IPCC）报告，全球气候变化预计将加剧极端火灾事件风险，野火发生频率净增长幅度可能达到约 30%30 \% （IPCC，2021）。近几十年来，加拿大、俄罗斯、亚马逊地区和澳大利亚南部等地发生的严重火灾事件已造成大范围森林损毁、人员伤亡和经济损失。
(Jain et al., 2024). In China, the average total annual area burned by forest fires was 0.4 Mha between 1987 and 2007 (Chang et al., 2015), affecting 78.6 % of the land (Chen et al., 2017). The period since the Last Glacial Maximum (LGM) includes major climatic shifts, such as deglaciation and the Holocene, which are key to understanding the relationships between fire activity, climate change, and vegetation dynamics. Studying wildfire history since the LGM helps clarify the natural variability of fire regimes and their drivers, supporting the interpretation of modern fire activity and predictions of future trends under climate change. （Jain 等，2024）。在中国，1987 年至 2007 年间森林火灾年均过火面积为 40 万公顷（Chang 等，2015），影响全国 78.6%的陆地面积（Chen 等，2017）。末次盛冰期（LGM）以来的时期包含了冰消期和全新世等重大气候变迁，这些时段对于理解火灾活动、气候变化与植被动态之间的关系至关重要。研究 LGM 以来的野火历史有助于厘清火情制度的自然变率及其驱动因素，为解释现代火灾活动及预测气候变化下的未来趋势提供支撑。
Eastern China’s climate is governed by the Asian monsoon system, and supports extensive primary productivity and hence fuel for fire (Ying et al., 2018; Zhang et al., 2015a). This region also houses about 96 华东地区的气候受亚洲季风系统控制，孕育了丰富的初级生产力，从而为火灾提供了充足的可燃物（Ying 等，2018；Zhang 等，2015a）。该区域还分布着约 96
% of the country’s population, and people have contributed to increased fire activity during the Holocene through accompanying agricultural development (Ma et al., 2018; Yuan et al., 2022; Zhang et al., 2021), and fire has, in turn, led to losses in human life and property. As a result, eastern monsoonal China is an excellent region for exploring the dynamics of fire and its relationship with different controlling factors. Prior research in China has tended to focus on fire histories from individual sites, including sites in northeast China (e.g. Li et al., 2017), on the Loess Plateau (Tan et al., 2015), on the Yungui Plateau (Zhang et al., 2015a) and in coastal regions (Ma et al., 2018; Wang et al., 2017). In addition, there are two studies that have attempted to synthesize records, to consider the broad-scale fire history in China, and its controls during the Holocene (Xu et al., 2021; Xue et al., 2018). Despite these researches, our understanding of long-term and spatiotemporal perspectives of fire activity at the regional scale and the responses of fire to climate change are in its infancy. 该国人口占比中，人类活动通过伴随的农业发展促进了全新世期间火灾活动的增加（Ma 等，2018；Yuan 等，2022；Zhang 等，2021），而火灾反过来又导致了人员生命和财产损失。因此，中国东部季风区是探索火灾动态及其与不同控制因素关系的理想区域。中国先前的研究往往集中于单个地点的火灾历史，包括中国东北地区（如 Li 等，2017）、黄土高原（Tan 等，2015）、云贵高原（Zhang 等，2015a）以及沿海地区（Ma 等，2018；Wang 等，2017）的站点。 此外，已有两项研究尝试整合记录，探讨中国全新世时期大范围火灾历史及其控制因素（徐等，2021；薛等，2018）。尽管存在这些研究，我们对区域尺度火灾活动的长期时空格局以及火灾对气候变化的响应机制仍处于初步认知阶段。
Two distinct fire responses regard to climatic conditions across China have emerged. Some studies suggest that warm and moist conditions favor more fires due to increased productivity, accumulation of biomass and fuel availability in north China. For example, five fire records from the Loess Plateau indicate that burning intensified throughout the mid- 中国各地气候条件呈现出两种截然不同的火灾响应模式。部分研究表明，在华北地区，温暖湿润的环境会因生产力提升、生物量积累及可燃物增多而引发更多火灾。例如，黄土高原的五份火灾记录显示，整个中全新世时期火势呈现加剧趋势。
Holocene, attributed to prevalent steppe cover, grass biofuel loads, and high precipitation and temperature (Wang et al., 2023). Similarly, in the Gonghai Lake from north China, where warm-humid climatic conditions created favorable environments for biomass accumulation, consequently enhancing wildfire likelihood which is correlated with intensified monsoon activity (Ji et al., 2021). In contrast, other studies have argued that dry climatic conditions are more conducive to fire occurrences. As an example, a soot record from Qinghai Lake suggested heightened wildfire activity during abrupt dry climate events (Hao et al., 2023). Additionally, a macroscopic charcoal record from Tengchong Qinghai Lake also described that enhanced fire activity corresponded with cold and dry climatic conditions, responding to the weak summer monsoon influence (Xiao et al., 2017). These contrasting observations highlight dry/wet climatic shifts in regulating fuel availability and combustibility, thereby influencing wildfire occurrences. Monsoon variability sets the patterns of rainfall and intensity in eastern China, and is thus likely to be a key control of fuel production and length of periods where fires are likely to occur. However, more composite evidence is required to better evaluate the overall contributions of climate and vegetation to fire occurrence in China. 全新世时期，草原植被广泛分布、草本生物燃料充足以及较高的降水与温度共同促进了野火活动（Wang 等，2023）。类似地，在中国北部的公海湖，温暖湿润的气候条件为生物量积累创造了有利环境，从而增加了与季风活动增强相关的野火发生概率（Ji 等，2021）。相反，其他研究认为干燥气候条件更易引发火灾。例如，青海湖的烟炱记录表明，在突发性干旱气候事件期间野火活动加剧（Hao 等，2023）。此外，腾冲青海湖的宏观木炭记录也显示，火灾活动增强与寒冷干燥的气候条件相对应，这响应了夏季风影响的减弱（Xiao 等，2017）。 这些对比鲜明的观测结果凸显了干湿气候转变在调节燃料可用性和可燃性方面的重要作用，从而影响野火发生。季风变化决定了中国东部的降雨模式和强度，因此很可能是燃料生产周期和火灾易发时段长度的关键控制因素。然而，需要更多综合性证据来更好地评估气候和植被对中国火灾发生的整体影响。
Here, we compiled 70 fire records, including charcoal and black 这里我们整理了 70 份火灾记录，包括木炭和黑色
Fig. 1. Fire records in this study. Bule dots record come from southern China and red dots come from northern China. Broad vegetation types provide the background (vegetation data were downloaded from https://www.resdc.cn/). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 图 1. 本研究中的火灾记录。蓝点代表中国南方记录，红点代表中国北方记录。背景为广义植被类型（植被数据下载自 https://www.resdc.cn/）。（关于图例中颜色引用的解释，请参阅本文的网络版。）
carbon proxies, from monsoon China covering the period since the LGM. The study aimed to explore the spatial characteristics of fire activity by synthesizing fire at regional scales and to consider the interactions between fire, climate and vegetation changes; hence enhancing the understanding of wildfire across different regions and to provide insights into future fire management strategies in the context of global warming. 末次盛冰期以来中国季风区的碳代用指标研究。该研究旨在通过综合区域尺度的火灾活动，探讨火灾活动的空间特征，并考虑火灾、气候与植被变化之间的相互作用；从而增进对不同地区野火现象的理解，并为全球变暖背景下的未来火灾管理策略提供见解。
2. Materials and methods 2. 材料与方法
2.1. Study region 2.1. 研究区域
This study focuses on the eastern monsoon-influenced region of China, including the transitional monsoon zone. We divided this region into two subregions-southern China (SC) and northern China (NC) based on vegetation types (vegetation data were obtained from Resource and Environmental Science Data Platform, https://www.resdc.cn/), as biomass and fuel composition directly influence fire activity. The boundary between the subtropical evergreen broad-leaved forest and the temperate deciduous broad-leaved forest aligns with the mean annual 800 mm precipitation line and the Qinling Mountain-Huai River boundary (Fig. 1). One of the key climatic characteristics of this region is the alternation between dry and wet seasons, coupled with the seasonal reversal of atmospheric circulation under the influence of the Asian monsoon. The mean annual precipitation (MAP) ranges from 800 to 1600 mm in SC and from 200 to 800 mm in NC. The distribution of vegetation types is illustrated in Fig. 1. 本研究聚焦于中国东部受季风影响的区域，包括季风过渡带。根据植被类型（植被数据来源于资源环境科学与数据中心平台，https://www.resdc.cn/），我们将该区域划分为两个子区域——华南地区（SC）和华北地区（NC），因为生物量和燃料组成直接影响火灾活动。亚热带常绿阔叶林与温带落叶阔叶林的分界线与年均 800 毫米降水线及秦岭-淮河界线相吻合（图 1）。该区域的关键气候特征之一是干湿季交替，同时受亚洲季风影响下大气环流的季节性反转。 南方地区年均降水量为 800 至 1600 毫米，北方地区则为 200 至 800 毫米。植被类型分布如图 1 所示。
2.2. Data sources 2.2. 数据来源
To investigate the fire history in the eastern monsoon region of China, we compiled charcoal and black carbon records from publications across the region, adopting these criteria: (1) The sites are located within the monsoon zone, excluding records from the Tibetan Plateau; (2) The records contain a chronology with at least three age control points and; (3) The records have no evidence of depositional hiatus in their sediments; and (4) The sampling resolution should be at least 1000 years per sample. For sites where raw data was not available, we digitized the fire proxy records from the original articles. Where the original figures represent charcoal or black carbon data changing with depth, these were converted them into a chronological sequence using their age-depth model with R package ‘Bacon’. 为研究中国东部季风区的火灾历史，我们从该地区已发表文献中整理了木炭和黑碳记录，采用以下筛选标准：（1）研究点位位于季风区范围内，排除青藏高原的记录；（2）记录需包含至少三个年代控制点的年代框架；（3）沉积物中无沉积间断的证据；（4）采样分辨率至少为每 1000 年一个样本。对于无法获取原始数据的点位，我们从原始论文中对火灾代用指标记录进行了数字化处理。当原始图表显示木炭或黑碳数据随深度变化时，我们使用 R 软件包'Bacon'中的年代-深度模型将其转换为时间序列。
2.3. Data treatment 2.3. 数据处理
Given that fire records were obtained using various techniques and units, such as percentages, concentration or accumulation rates, a standardized method was applied to ensure comparability across sites over time. The data treatment protocol consisted of the following steps, based on methods described by Power et al. (2008): 鉴于火灾记录是通过多种技术和单位（如百分比、浓度或累积率）获取的，我们采用标准化方法以确保不同地点随时间变化的数据可比性。数据处理方案基于 Power 等人（2008 年）描述的方法，包含以下步骤：
(1) Minimax transformation: Convert the data using the formula c_(i)\mathrm{c}_{\mathrm{i}}, =(c_(i)-c_("min "))//(c_("max ")-c_("min "))=\left(c_{i}-c_{\text {min }}\right) /\left(c_{\text {max }}-c_{\text {min }}\right). (1) 极小极大变换：使用公式 c_(i)\mathrm{c}_{\mathrm{i}} 、 =(c_(i)-c_("min "))//(c_("max ")-c_("min "))=\left(c_{i}-c_{\text {min }}\right) /\left(c_{\text {max }}-c_{\text {min }}\right) 对数据进行转换。
(2) Log transformation: Apply c_(i)^(**)=log(c_(i)^(**)+0.01)\mathrm{c}_{\mathrm{i}}{ }^{*}=\log \left(\mathrm{c}_{\mathrm{i}}{ }^{*}+0.01\right), the addition of 0.01 was used to avoid undefined values for zero data points. (2) 对数变换：应用 c_(i)^(**)=log(c_(i)^(**)+0.01)\mathrm{c}_{\mathrm{i}}{ }^{*}=\log \left(\mathrm{c}_{\mathrm{i}}{ }^{*}+0.01\right) ，添加 0.01 以避免零数据点出现未定义值。
(3) Z-scores calculation: Calculate Z-scores as z=(c_(i)^(**)^(-)^(-)c_(i)^(**))//sigma\mathrm{z}=\left(\mathrm{c}_{\mathrm{i}}{ }^{*}{ }^{-}{ }^{-} \mathrm{c}_{\mathrm{i}}{ }^{*}\right) / \sigma. (3) Z 值计算：按 z=(c_(i)^(**)^(-)^(-)c_(i)^(**))//sigma\mathrm{z}=\left(\mathrm{c}_{\mathrm{i}}{ }^{*}{ }^{-}{ }^{-} \mathrm{c}_{\mathrm{i}}{ }^{*}\right) / \sigma 公式计算 Z 值。
Thus, all resulting data are dimensionless. The Z-scores were then linearly interpolated Z-scores within a 100-year window for sites with chronology resolutions finer than 100-200 years per sample and for sites with coarser resolutions, a 500-year interpolation window was applied. The interpolated Z-scores were then composited into regional fire reconstructions using the locally weighted smoothing (LOWESS). 因此，所有生成的数据均为无量纲值。对于年代分辨率高于每样本 100-200 年的站点，采用 100 年窗口进行 Z 分数的线性插值；对于分辨率较低的站点，则应用 500 年插值窗口。随后通过局部加权平滑法（LOWESS）将插值后的 Z 分数整合为区域火灾重建序列。
3. Results 3. 结果
A total of 70 charcoal and black carbon records were identified for 共识别出 70 份木炭和黑碳记录
this study, including 37 from SC and 33 from NC (see details in SI), covering the past 21 ka BP . The synthesized wildfire history shows that there is an asynchronous pattern between the southern (SC) and northern (NC) regions, indicating significant spatial variability (Fig. 2). The southern region exhibited higher fire activity while the northern region experienced reduced fire incidents. 本研究共收集了 37 份来自南方地区（SC）和 33 份来自北方地区（NC）的样本（详见附录），时间跨度覆盖过去 2.1 万年。综合野火历史分析显示，南方（SC）与北方（NC）区域存在异步变化模式，表明具有显著的空间异质性（图 2）。南方区域呈现出更高的火灾活动频率，而北方区域则经历了火灾事件的减少。
During the Last Glacial Maximum (LGM), the results revealed that fire in SC was relatively frequent, but then dramatically decreased after 17.5 ka BP. NC exhibited predominantly negative Z scores from 21 to 15 ka BP, indicating lower wildfire activity. The lowest period in fire activity for SC occurred during the Bølling-Allerød interstadial (B/A, 14.8-12.9kaBP14.8-12.9 \mathrm{ka} \mathrm{BP} ), followed by an increase around the Younger Dryas event (YD, 12.8-11.6 ka BP). In contrast, NC fire activity displayed a moderate increase during the B/A and a decrease during the YD. In SC, relatively low fire activity was evident in the mid-Holocene, followed by a gradual decrease during the late Holocene. Conversely, NC fire activity showed a rising trend throughout the Holocene, peaking between 7 and 6 ka BP and again in the last 1 ka BP . 末次盛冰期（LGM）期间，研究结果显示南方地区（SC）火灾发生频率相对较高，但在距今 17.5 千年后显著减少。北方地区（NC）在 21 至 15 千年间主要呈现负 Z 值，表明野火活动较弱。南方地区火灾活动最低值出现在波令-阿勒罗德间冰阶（B/A， 14.8-12.9kaBP14.8-12.9 \mathrm{ka} \mathrm{BP} ），随后在新仙女木事件（YD，12.8-11.6 千年）前后有所增加。与之相反，北方地区火灾活动在 B/A 期间适度增强，而在 YD 期间减弱。南方地区在中全新世火灾活动明显偏低，随后在晚全新世逐渐减少。而北方地区在整个全新世呈现上升趋势，分别在 7-6 千年间和最近 1 千年达到峰值。
4. Discussion 4. 讨论
4.1. Regional fire evolution connected to fuels 4.1. 区域火势演变与可燃物的关联
The occurrence of wildfire requires the triangular intersection of fuel load, heat source and conducive meteorological conditions (Moritz et al., 2005; Whitlock et al., 2010). Fuel consists of various organic materials, including wood, grasses, leaves, and other plant matter, and this forms the foundational basis to carry a fire (Keane, 2013; Labenski et al., 2022). Combustible matters need to reach ignition point to burn, the natural ignition source is lightning, but human induced fire became prevalent in the Holocene (Pei et al., 2020). Meteorological and antecedent conditions, such as high temperature, low moisture and strong wind significantly affect the emergence and diffusion of fires. Among these factors, fuel loads are the most important control of regional fires under natural condition in the Asian summer monsoon region (Ji et al., 2021). The density and type of vegetation influences the spread and intensity of wildfires, the areas with abundant dry fuels such as grasses and shrubs can ignite more easily and spread more rapidly compared to with sparse or moist vegetation. For example, broadleaf-dominated forests, particularly those with Alnus, Lithocarpus/Castanopsis and tropical tree species, demonstrate considerable fire resistance due to their high moisture retention capacity and low flammability leaf litter (Xiao et al., 2017). Dry conifer forests are particularly vulnerable to the combined effects of altered fire regimes and climate change (McClure et al., 2024). In this study, we connected the composite regional fire 野火的发生需要燃料负荷、热源和有利气象条件三要素的交集（Moritz 等，2005；Whitlock 等，2010）。燃料由各种有机物质组成，包括木材、草类、树叶和其他植物材料，这些构成了火灾蔓延的物质基础（Keane，2013；Labenski 等，2022）。可燃物需达到燃点才能燃烧，自然火源主要是闪电，但全新世以来人为火源逐渐盛行（Pei 等，2020）。高温、低湿和强风等气象及前期条件会显著影响火灾的发生与扩散。在亚洲夏季风区自然条件下，这些因素中燃料负荷是区域火灾最重要的控制因素（Ji 等，2021）。 植被的密度和类型会影响野火的蔓延速度和强度。与植被稀疏或湿润的地区相比，拥有丰富干燥燃料（如草类和灌木）的区域更容易被点燃且火势蔓延更快。例如，以阔叶树为主的森林（尤其是含赤杨属、石栎/栲属及热带树种的林区）因其强大的保水能力和低可燃性的凋落物层而表现出显著的抗火性（Xiao 等，2017）。干燥的针叶林尤其容易受到火灾模式改变和气候变化的双重影响（McClure 等，2024）。本研究中，我们将复合区域火灾...
Fig. 2. Composite fire history in southern and northern China. The dotted line indicates a 95 % confidence interval. 图 2. 中国南北方综合火灾历史记录。虚线表示 95%置信区间。
This article is part of a Special issue entitled: ‘Fire’ published in Global and Planetary Change. 本文是《全球与行星变化》期刊特刊"火灾"专题的一部分。
Corresponding author at: State Key Laboratory of Loess Science, Institute of Earth Environment, Chinese Academy of Sciences, Xi’an, Shaanxi 710061, China. 通讯作者：中国科学院地球环境研究所黄土与第四纪地质国家重点实验室，陕西西安 710061。
