The increased effect of spring leaf unfolding on autumn senescence in the northern and southern hemispheres

doi:10.21203/rs.3.rs-5190727/v1

Download PDF

Article

The increased effect of spring leaf unfolding on autumn senescence in the northern and southern hemispheres

https://doi.org/10.21203/rs.3.rs-5190727/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Plant phenology, the timing of periodic events in plant development, is an indicator for detecting climate-vegetation dynamics. Although the importance of vegetation growth carryover (VGC) on phenology was recognized in the Northern Hemisphere (NH), it is unclear how VGC and climatic factors contribute to phenology and how these contributions evolve at a global scale. Utilizing two sets of satellite NDVI data, we explored the impacts of climate change and VGC on start-of-season (SOS) and end-of-season (EOS) plant phenology during the past four decades. Here we show that, globally, advanced SOS resulted primarily from the increasing temperature and radiation, whereas delayed EOS was mainly attributed to the increasing temperature and VGC in SOS (VGC_SOS). VGC_SOS was the main driver of EOS in the southern hemisphere (SH), whereas it was temperature in the NH. Furthermore, the contribution of VGC_SOS to EOS displayed increasing trends over the past four decades in both hemispheres, which was particularly significant in NH. These differences were due primarily to the more evident "warming and drying" trends in the SH than NH, which weaken VGC_SOS's impact on vegetation EOS. We conclude that VGC_SOS contributes substantially to EOS in both hemispheres, offering insights for accurate predictions of vegetation growth and carbon sequestration under future global warming scenarios.

Earth and environmental sciences/Ecology/Climate-change ecology/Phenology

Earth and environmental sciences/Ecology/Ecosystem ecology

climate change

spring phenology

autumn phenology

vegetation growth carryover

Northern Hemisphere

Southern Hemisphere

Plant phenology is the annual cyclical phenomenon during the process of plant growth and development and enables plants to adapt to seasonal climatic and environmental changes (White et al. 2009). It encompasses stages such as germination, branching, leaf unfolding, flowering, and leaf fall (Schwartz et al. 2006; Piao et al. 2019). The start-of-season (SOS) and end-of-season (EOS) mark critical junctures in the plant growth cycle. Both SOS and EOS are sensitive to climate change and are pivotal in determining the vegetation growing season's duration, and its impact on ecosystem carbon cycles (Zhou et al. 2023). Given the critical roles of SOS and EOS in shaping the dynamics of terrestrial ecosystems, a deeper exploration of phenological variations and their driving factors across different global contexts is warranted for enhancing our understanding of how climate change interacts with terrestrial ecosystems.

Understanding the mechanisms behind changes in vegetation phenology would enable the assessing of ongoing impacts of climate change and the ecological consequences. Temperature is widely recognized as the key determinant of vegetation phenology (Fu et al. 2015; Hong et al. 2022), with global warming advancing spring leaf unfolding and delaying autumn senescence (Fu et al. 2015; Chen et al. 2018). However, winter warming may also extend the chilling period required by plants and potentially delay leaf unfolding radiation (Rossi & Isabel 2017; Guo et al. 2023). Radiation intensity and growth season length not only serve as consistent indicators for seasonal temperature shifts (Flynn & Wolkovich 2018), but they also directly trigger physiological morphology and even gene regulation in plants, thereby influencing both spring and autumn phenological events (Gill et al. 2015; Vitasse et al. 2018). Studies by Descals et al. (2023) reported that radiation facilitates the onset of spring phenology by providing the essential energy required for photosynthesis, which must coincide with appropriate temperature thresholds to effectively trigger photosynthetic activity in plants (Descals et al. 2023). Furthermore, Xie et al. (2015) concluded that autumn phenology responds intricately to environmental stressors, such as drought and precipitation (Xie et al. 2015), suggesting that plant growth and development respond in complex, non-linear ways to environmental changes (Peng et al. 2019; Wu et al. 2022). However, most research has focused predominantly on the NH, with limited studies on the drivers of vegetation phenology in the SH. Due to differences in biodiversity and climatic conditions between the two hemispheres, it would be of interest to examine and compare the impacts of different factors on vegetation phenology between them.

Vegetation phenology is shaped not only by climate change, but also by the preceding phenological state, known as vegetation growth carryover (VGC) (Lian et al. 2021). It is generally accepted that climatic factors are primary drivers of seasonal-to-interannual dynamics of vegetation growth and associated carbon uptake in the northern hemisphere (NH). However, the interconnectedness of vegetation growth across seasons is increasingly recognized. Notably, spring phenology influences the subsequent summer and autumn stages, underscoring the crucial role of spring vegetation in shaping seasonal development and community dynamics (Mei et al. 2021; Zeng et al. 2021). Specifically, an earlier SOS could lead to an earlier peak-of-season (POS) and EOS (Lian et al. 2021; Peng et al. 2021; Huang et al. 2023), and even to an earlier SOS in the following year (Fu et al. 2014). A further study concluded that these carryover effects from the previous season's growth surpassed the impacts of temperature and precipitation (Lian et al. 2021). However, solar radiation and other climatic factors were not considered during this comparison, which causes some uncertainty in the conclusions. Furthermore, the carryover effects received considerable attention across the NH, such as for temperate grasslands and forests, but it is remains uncertain as to whether this VGC also exists in the southern hemisphere (SH). To fill this gap necessitates identifying the spatial pattern of VGC, the relative contributions of climate change and VGC, their distribution among different vegetation types, and the comparison between the two hemispheres. In addition, plants synchronize their life activity cycles with seasonal climate changes, and they continually adapt to the changing climates (Li et al. 2023). Therefore, detecting changes in the contributions of climate and VGC to phenology over time could improve our understanding of vegetation dynamics under global climate change.

In the present study, we extracted SOS and EOS from two sets of NDVI datasets (GIMMS and MODIS) spanning 1982 to 2022, and determined the impacts of climate change (temperature, precipitation, radiation, and PET) and VGC on vegetation phenology at the global scale. Our main objectives were to: 1) compare the accuracy of GIMMS and MODIS datasets for SOS and EOS, as well as their spatio-temporal distribution patterns; 2) assess the effects of climate change and VGC on plant phenology; 3) analyze the spatio-temporal trends in the contributions of key climatic factors and VGC to phenological events over the last four decades; and 4) identify the differences in these impacts between the NH and SH and across different vegetation types. By addressing these four objectives, our study can provide novel insights into the mechanisms underlying vegetation growth changes.

Spatio-temporal patterns in phenology from GIMMS and MODIS datasets

Figure 1 illustrates the spatial distribution of annual mean SOS and EOS from GIMMS and MODIS datasets, respectively. Both datasets exhibited consistent spatial patterns of SOS and EOS (Fig. 1c, h). For SOS, SOS_GIMMS typically occurred at day of year (DOY) 90.64, while SOS_MODIS occurred at DOY 85.15, with both events occurring earlier (DOY 82.44) near the equator. Moreover, in areas north of 30° in the NH, SOS was delayed as the latitude increased. EOS derived from the MODIS dataset exhibited similar latitudinal and spatial patterns as the GIMMS dataset. EOS_GIMMS occurred at DO 265.83, while EOS_MODIS occurred at DOY 261.76, with both events occurring later (DOY 291.37) near the equator. Notably, the length of the growing season derived from both datasets was 176.61 days. Overall, the SOS and EOS from the two datasets displayed similar spatial patterns in the NH (RMSE of 9.56 and 12.73, respectively), but the spatial variability in the SH was much larger (RMSE of 33.88 and 35.89, respectively) (Figs. 1d-e and 1i-j). Moreover, the temporal changes of the phenology extracted from the two datasets differed (Supplementary Fig. 3e, g).

SOS_GIMMS and SOS_MODIS generally displayed similar latitudinal and spatial patterns in their temporal variations (Supplementary Fig. 2). Globally, the percentage of pixels with an advanced SOS consistently reached 60.97% for MODIS and 59.93% for GIMMS NDVI3g data. In the NH, SOS displayed an advancing trend in 64.85% of the pixels for GIMMS data and in 58.28% for MODIS data. The SH displayed a higher percentage of pixels with a delayed SOS in GIMMS data (56.07%) compared to those in MODIS data (49.76%). Specifically, SOS advancements in the NH were located primarily in central and western Siberia, central North America, and northeastern China, while delays in the SH were concentrated largely in the southern Amazon and central Africa. The temporal trend in EOS extracted from both datasets exhibited spatial consistency, with less pixels showing significant changes in EOS for MODIS data (Supplementary Figs. 2c-d). In the NH, delayed EOS was noted in larger areas for both datasets (52.21% for GIMMS and 58.39% for MODIS), and occurred primarily in eastern North America, Europe, central and western Asia, and northern Africa. In contrast, the SH's EOS_GIMMS exhibited a much greater percentage of delayed pixels (62.08%) compared to EOS_MODIS (50.94%), and were distributed primarily in central and southern North America, Europe, central and eastern Asia, and central Africa. In addition, SOS exhibited a significant advancing trend in both datasets, averaging 0.1364 days/year for GIMMS (1982–2015) and 0.1051 days/year for MODIS (2001–2022) (Supplementary Fig. 3). Although EOS displayed a similar delaying trend in both datasets, 0.1209 days/year for GIMMS and 0.1179 days/year for MODIS, there was significant divergence in temporal changes between 2010 and 2015 for the two datasets.

The relationships between phenology and climatic factors

Supplementary Fig. 4 presents the spatial patterns of the partial correlation coefficients between SOS derived from the two datasets and climatic factors (temperature, precipitation, radiation, and PET) at the 3-month scale. The relationship between SOS and temperature revealed a similar spatial distribution for both datasets, with a negative correlation between temperature and SOS in most regions (SOS_GIMMS: 70.89%, SOS_MODIS: 82.71%) (Supplementary Figs. 4a, f). Additionally, 15.43% and 17.65% of the regions in the SOS_GIMMS and SOS_MODIS studies, respectively, correlated negatively (P < 0.05) with temperature. There was a negative correlation between SOS and precipitation (SOS_GIMMS: 51.01% and SOS_MODIS: 61.47%) in over half the regions (Supplementary Figs. 4b, g). The correlation between SOS and radiation was negative in approximately 60% of the areas (SOS_GIMMS: 68.13% and SOS_MODIS: 60.99%), with significant changes in 8.37% and 4.98% of the pixels (P < 0.05), mainly in Europe, parts of arid Asia, and southern Australia (Supplementary Figs. 4c, h). However, the proportions of positive and negative correlations between SOS and PET were roughly equal for both datasets, with neither dominating the study area (Supplementary Fig. 4d, i). In the correlation between VGC_EOS and SOS, more than half of the regions were correlated negatively (SOS_GIMMS: 58.76%; SOS_MODIS: 55.16%), and the regions with significant changes were relatively small (Supplementary Fig. 4e, j).

EOS was correlated positively with temperature in more than 80% of the areas (EOS_GIMMS: 82.15% and EOS_MODIS: 80.40%), with 10.39% and 15.41%, respectively, of the pixels exhibiting significant positive correlations (P < 0.05). These positive correlations were distributed primarily in high-latitude regions and near the equator (Fig. 2a, f). The influence of precipitation on EOS was also widespread across many areas. The EOS_GIMMS was correlated positively with precipitation in 78.83% of the pixels, with significant positive correlations accounting for 20.75% (P < 0.05), and distributed mainly in arid regions worldwide. For EOS_MODIS, the percentage of positive correlations with precipitation was 80.97%, with significant positive correlations accounting for 13.42% (P < 0.05), distributed mainly in the NH, but also sporadically in the SH. (Fig. 2b, g). Compared to temperature and precipitation, radiation had a smaller impact on EOS, with the percentages of positively correlated pixels for EOS_GIMMS and EOS_MODIS being 53.16% and 51.31%, respectively, and of pixels with significant changes being 4.38% and 4.46% (P < 0.05) (Fig. 2c, h). The positive (49.78%) and negative (50.22%) correlations between PET and EOS_GIMMS were roughly equal, with neither dominating (Fig. 2d, i). In addition, temperature, radiation, and PET at the 3-month scale for both SOS and EOS displayed significant increases during 1982–2022, except for precipitation (Supplementary Figs. 6–8). VGC_SOS and EOS were correlated positively in most regions of the world (GIMMS: 81.32%; MODIS: 83.14%), with significant correlation in 22.04% and 25.33% (P < 0.05) of the areas, respectively, and distributed mainly in North America, Europe and Central Asia. Only a few regions showed negative correlations (GIMMS: 18.68%; MODIS: 16.86%), with significant negative correlations in 7.38% and 6.12% (P < 0.05) of the areas, respectively (Fig. 2e, j).

The contribution of climatic factors and VGC to phenology

The contributions of climatic factors and VGC to phenology differed between the two hemispheres (Fig. 3). In the NH, temperature had the largest contribution to advancing SOS at -0.19 days/year, followed by radiation at -0.08 days/year. PET delayed the occurrence of SOS by 0.04 days/year, while VGC_EOS had the smallest contribution (Fig. 3a). In the SH, radiation was the most influential factor on SOS at -0.13 days/year, with temperature following closely. Both precipitation and PET delayed SOS, and VGC_EOS also had a significant contribution at -0.05 days/year (Fig. 3b). For EOS, temperature (0.19 days/year) was the most influential climatic factor in the NH, followed by VGC_SOS (0.15 days/year) (Fig. 3c). A similar trend was observed in the SH, where the contribution of VGC_SOS to EOS (0.18 days/year) was greater than temperature (0.13 days/year) and other factors (Fig. 3d).

Figure 4 illustrates the contributions of climatic factors and VGC_SOS to EOS for various vegetation types in both hemispheres. Overall, in the NH, temperature had the greatest contribution to the EOS for all vegetation types, followed by VGC_SOS and precipitation. In the SH, all factors, except PET, contributed significantly to delaying EOS, depending on vegetation types. For forests, both VGC_SOS (0.15 days/year) and radiation (0.16 days/year) contributed significantly to the EOS in the SH (Fig. 4e). For both shrublands and savannas, the contribution of temperature was greater than VGC_SOS (Fig. 4f, g), and for grasslands, temperature (0.19 days/year) and precipitation (0.18 days/year) had similar influences on EOS (Fig. 4e-h).

Supplementary Fig. 9 illustrates the contributions of climatic factors and VGC_EOS to SOS for various vegetation types in both hemispheres. Overall, temperature was the largest contributor to the SOS for all vegetation types in the NH, indicating that warming advances SOS. However, the critical factors influencing SOS in the SH differed among vegetation types, with radiation, precipitation, and temperature all playing key roles. For forests, in the NH, temperature was the most important contributor to the SOS trend (-0.14 days/year), followed by radiation (-0.06 days/year) (Supplementary Fig. 9a). In the SH, radiation had the greatest influence, advancing the SOS by 0.26 days/years, while precipitation delayed the SOS of forests (Supplementary Fig. 9c). The remaining factors, including PET and VGC_EOS, tended to delay SOS for forests in both hemispheres. For shrublands, temperature was also the greatest contributor to SOS in the NH (-0.27 days/year), significantly greater than the combined contributions of the other four factors (Supplementary Fig. 9b). In the SH, precipitation had the greatest impact on SOS (-0.11 days/year), followed by temperature (-0.05 days/year) and radiation (-0.01 days/year) (Supplementary Fig. 9f). For savannas, like shrublands, precipitation contributed most in the SH (Supplementary Fig. 9c, g). In grasslands, temperature was the dominant contributor to SOS in both hemispheres, -0.18 days/year in NH and − 0.11 days/year in SH. In the SH, contribution of precipitation (-0.09 days/year) was comparable to temperature (Supplementary Fig. 9h).

The spatial distribution of relative contributions of VGC and climatic factors to phenology is presented in Fig. 5. For SOS, temperature (28.08%) and radiation (23.45%) were the primary influences on a global scale, while VGC_EOS was relatively weak (14.12%) (Fig. 5b). Specifically, in the NH, temperature (36.32%) and radiation (26.89%) remained the main factors affecting SOS, while VGC_EOS had the smallest impact (5.64%). In the SH, radiation (41.33%) was the dominant factor for SOS, with PET (24.00%) and VGC_EOS (22.41%) having greater influences than temperature and precipitation. Spatially, high-latitude areas in the NH were influenced mainly by temperature, while radiation dominated in central North America, Europe, and regions near the equator (Fig. 5a).

Globally, EOS was influenced primarily by temperature (28.87%) and VGC_SOS (27.05%), followed by precipitation (19.09%) (Fig. 5d). The NH's EOS was similar to the global EOS, being influenced mainly by temperature (37.58%), VGC_SOS (25.64%), and precipitation (22.18%). In contrast, in the SH, EOS was affected predominantly by VGC_SOS, which influenced 66.53% of the region. Following VGC_SOS, temperature (10.52%), radiation (9.55%) and precipitation (7.93%) also contributed to the EOS in the SH. Spatially, temperature and VGC_SOS were the main factors influencing high-latitude regions, radiation dominated near the equator, and VGC_SOS was the main factor in the South-central Amazon, southern Africa, and northern Australia (Fig. 5c).

The spatio-temporal variation characteristics of the contribution of major climatic factors and VGC

The temporal variation in the contribution (Tcon) of VGC to phenology with time was characterized primarily by significant changes (P < 0.05) in pixels, including increases (Tcon+) and decreases (Tcon-) (Fig. 6). For the contribution of VGC_EOS to SOS, nearly half of the pixels displayed significant changes in Tcon globally. In the NH, the contribution of VGC_EOS exhibited an increasing trend across a series of moving windows (slope = 0.005, P < 0.01), shifting from advancing SOS in the first moving window to delaying it in the last moving window (Fig. 6b). Specifically, the percentage of pixels with significant increases in the contribution of VGC_EOS (24.47%) was slightly higher than those with significant decreases (21.99%) (Fig. 6a). In contrast, in the SH, the overall temporal trend increased slightly (slope = 0.001, P < 0.05) over time (Fig. 6c). Spatially, the significant decrease in the contribution of VGC_EOS to SOS occupied a higher percentage of pixels (23.03%) than those with a significant increase (21.40%) (Fig. 6a).

For the contribution of VGC_SOS to EOS, both hemispheres had significant increases over the past four decades, with larger increases in the SH (slope = 0.005) than NH (slope = 0.001) (Fig. 6e, f). The significant increases were distributed primarily around the equator, eastern and central North America, Europe, and central Asia (Fig. 6d). In the NH, the significant increases in the contributions of VGC_SOS accounted for 26.97%, nearly twice the percentage of pixels displaying significant decreases (13.41%). In the SH, the significant increases and decreases were comparable, 17.70% and 15.24%, respectively (Fig. 6d).

We also quantified the temporal variations in the contribution of major climatic factors (temperature, precipitation, and radiation) to SOS and EOS (Supplementary Fig. 12). Temporally, in the NH, an increase in temperature (slope = -0.004, P < 0.01) and radiation (slope = -0.005, P < 0.01) generally decreased, while precipitation generally increased (slope = 0.001, P < 0.05) EOS. However, the impact of climatic factors varied substantially among regions and hemispheres.

The changes in the contribution of VGC to SOS and EOS among vegetation types are depicted in Fig. 7. Overall, VGC_EOS tended to advance SOS during the past 40 years for all vegetation types, with an average of -0.007 days/year (Fig. 7a). However, the advancement in SOS due to VGC_EOS decreased from − 0.012 days/year during 1982–2000 to -0.004 days/year during 2004–2022 (Fig. 7a, b). Savannas underwent a similar change in the effect of VGC_EOS on SOS, where the advancement in SOS decreased from − 0.017 days/year to -0.003 days/year over the same two periods. The greatest difference between the two periods in the effect of VGC_EOS on SOS occurred in forests (0.023 days/year), while the smallest difference occurred in grasslands (0.01 days/year). For both forests and grasslands, the VGC_EOS shifted SOS from advancing to delaying it. In contrast, VGC_EOS had an opposite effect on SOS for shrublands, shifting from delaying SOS to advancing it (Figs. 7a, c and e). Moreover, the contribution of VGC_SOS to EOS exhibited an increasing trend over the two periods (P < 0.05), with different vegetation types exhibiting varying degrees of increase (Fig. 7d, f); the greatest increase (0.031 days/year) occurred in savannas between the two periods and the smallest increase (0.024 days/year) occurred in both forests and grasslands (Fig. 7b).

The changes in contributions of major climatic factors to SOS and EOS over time for different vegetation types are illustrated in Supplementary Figs. 13 and 14, respectively. The absolute contribution of precipitation to advancing SOS decreased from 1982–2000 to 2004–2022 for savannas (Supplementary Fig. 13b, g). Similarly, both shrublands and grasslands exhibited significant decreases in the absolute contribution of radiation to SOS (Supplementary Fig. 13c, g ). In contrast, for savannas, the contributions of precipitation and radiation to EOS increased significantly over the two periods (Supplementary Fig. 14g).

Phenological data extracted from two datasets

We extracted two sets of long-term phenological data from the GIMMS (19982 − 2015) and MODIS (2001–2022) NDVI datasets for the whole globe. Both datasets exhibited spatio-temporal consistency in phenology, marked by an advancement of SOS and a delay in EOS. However, the extracted phenology (SOS and EOS) from MODIS occurred approximately 4.78 days earlier than from GIMMS. Consistent with a recent study (Descals et al. 2023), we found a greater advancement in SOS than delay in EOS. In the NH, SOS displayed an overall advancing trend for both GIMMS and MODIS datasets, while in the SH, the percentage of pixels with delayed SOS_GIMMS (56.07%) was greater than those with advanced SOS_GIMMS. But the percentages of delayed and advanced SOS_MODIS pixels were similar for MODIS data in the SH. Similarly, both datasets presented an over delayed trend in EOS in the NH, while the percentages of delayed EOS in the SH were 62.08% for GIMMS and 50.94% for MODIS. Moreover, after 2010, EOS_GIMMS displayed a sharp decline, while EOS_MODIS displayed a slow upward trend. Previous studies suggested that post-2000, GIMMS_NDVI and MODIS_NDVI exhibited divergent phenological trends, attributable largely to differences in data acquisition methods, sensor types, and data processing and analysis (Ge et al. 2015; Wang et al. 2024). As we applied the same NDVI processing and phenological extraction procedures to both datasets, we reason that the differences in phenological extraction results could be due to different data sources and regional vegetation differences. Firstly, GIMMS NDVI data originated from the older AVHRR sensor, while MODIS NDVI used the more advanced Terra sensor (An et al. 2014). These two sensors differ in design, spectral response, and calibration, particularly in their handling of cloud cover and atmospheric effects (Fensholt & Proud 2012). These differences can lead to substantial discrepancies between the two data sets in the SH (Fensholt et al. 2009). Moreover, cloud and atmospheric conditions in the SH may differ from those in the NH, which could also contribute to these discrepancies (Tarnavsky et al. 2008). Secondly, there are considerable differences in vegetation types and distribution between the SH and NH (Guay et al. 2014). The SH encompasses extensive tropical rainforests and grasslands, whose seasonal and growth responses differ from those of the temperate and boreal forests in the NH (Jiang et al. 2017a). These ecological differences can impact NDVI measurements since various vegetation types respond differently to light and temperature (Ye et al. 2021). It is worth noting that satellite-based NDVI data can introduce biases into the analysis because its definition of phenology differs from ground observations (Fensholt et al. 2009; Fensholt & Proud 2012). In the future, site data and field phenological observation data should be employed to correct the phenological information extracted from multi-source NDVI data.

The dominant roles of climatic factors and VGC in phenology

Phenology often exhibits lagged responses to climate change and VGC (Chang et al. 2019). The highest partial correlation coefficients between phenology and climatic factors emerged at the 3-month scale, indicating that short-term climate changes had strong impacts on phenology. Temperature and solar radiation were the main climatic drivers of SOS in the current study, which is in agreement with previous findings (Liu et al. 2016a). However, the dominant factors differed between the NH and SH; temperature was the primary driver of SOS in the NH, whereas it was radiation in the SH. The NH contains vast temperate and boreal forests that are highly sensitive to temperature variations (De Pue et al. 2023). Specifically, rising spring temperatures can rapidly trigger vegetation growth, rendering temperature as the key climatic factor affecting vegetation in the NH (Fensholt et al. 2009). Furthermore, water availability is linked frequently to temperature changes, reinforcing temperature's dominant influence on vegetation (Guay et al. 2014). In contrast, the SH is characterized by extensive tropical rainforests and savannas, where vegetation is dependent on radiation (Neimane-Sroma et al. 2024). In tropical regions, temperature remains relatively stable throughout the year, positioning radiation as the key limiting factor for vegetation productivity (Ye et al. 2021). Particularly in arid and semi-arid regions, radiation intensity directly affects the efficiency of photosynthesis, establishing radiation as the most important climatic factor affecting vegetation in the SH (Jiang et al. 2017a). Temperature and precipitation are the main climatic factors determining the occurrence of EOS in both the NH and SH. Precipitation directly affects the availability of water for plants, thereby influencing photosynthesis and water use efficiency (Xie et al. 2015). Additionally, the influence of climatic factors on phenology displayed significant spatio-temporal heterogeneity, with the impact mechanisms differing among geographic locations, ecosystem types, and the physiological characteristics of specific species (Liu et al. 2016a; Li et al. 2018; Nagai et al. 2020).

In contrast, lower partial correlation coefficients were observed between phenology and the climatic factors in the preceding 6 to 12 month periods, indicating a gradual diminishment in the influence of climatic factors on phenology with increasing time span. This can be attributed primarily to the plants’ intrinsic biological mechanisms and their adaptability to long-term environmental changes (Liu et al. 2023).

However, vegetation growth is a continuous process of change, influenced not only by climatic factors but also by VGC (Lian et al. 2021). In the present study, VGC_SOS had a much greater impact on the current year's EOS than VGC_EOS had on the next year's SOS, and this occurred in the majority of vegetation types. (Fig. 8, 9). For EOS, the contribution of VGC_SOS was slightly lower than that of temperature in the NH, while the EOS of vegetation in the SH was dominated by VGC_SOS. This was related primarily to climatic conditions, vegetation types, and their biogeographical distribution across different hemispheres (Zhou et al. 2021; Yuan et al. 2022). There were consistent patterns for all vegetation types in the NH. Specifically, temperature contributed the most to the delay of EOS, followed by VGC_SOS and precipitation. Notably, VGC_SOS had the greatest contribution to the delay of EOS for forests in the SH, greater than even temperature. For shrublands and savannas in the SH, the contribution of VGC_SOS was second only to temperature, while for grasslands in the SH, it was exceeded by both temperature and precipitation. According to previous studies, spring plays a decisive role in vegetation growth dynamics, biomass accumulation, and energy balance throughout the entire growing season (Chen et al. 2019). During this period, forests fix carbon and convert energy through photosynthesis efficiently, enabling productivity in the subsequent growing season (Zhao et al. 2022). The quality of spring growth conditions directly determines the growth potential and resilience of vegetation, thereby influencing EOS (Liu et al. 2018; Chen et al. 2022). In general, grasslands are the least affected by VGC_SOS among all vegetation types. This is because grassland ecosystems, due to their short lifecycle and rapid growth, are very sensitive to immediate climate conditions, especially temperature and precipitation (Liu et al. 2016a). As a result, the EOS of grasslands reflects the climatic conditions of the current year, unlike woody plants that are influenced by the effects of spring growth conditions over the long term (Tagesson et al. 2014). The smaller impact of VGC_EOS than VGC_SOS on the next year's SOS may be due to plants experiencing a period of physiological dormancy during winter. During this time, physiological activities decelerate, and accumulated effects are mitigated or reset (Nilsson 2022). Therefore, the onset of spring growth is more directly influenced by the climatic conditions of the current spring (Gillespie & Volaire 2017).

Spatio-temporal trend of the contributions of dominant factors to phenology

Using a moving window model, the current study determined the spatio-temporal variation trends in the contributions of major climatic factors (temperature, precipitation, and radiation) and VGC to SOS and EOS over the past 40 years. For SOS, the contribution by temperature and radiation in advancing SOS tended to decelerate globally, which is consistent with the findings of Fu et al (Fu et al. 2015). Temporally, the contribution of only radiation to SOS exhibited decreases in the NH, and the percentage of pixels displaying increases was still greater than displaying decreases; whereas the opposite was true in the SH. Additionally, the contribution of radiation to advancing SOS for shrublands and grasslands was also decreasing. This may be a result of the SH experiencing more cloud cover, leading to reduced radiation (Liu et al. 2016b). Conversely, the contribution of precipitation to SOS displayed a decreasing trend in both hemispheres, indicating that the delayed SOS caused by precipitation in earlier years may have shifted to advancing SOS during the last 15 years, probably due to reduced precipitation. This could be because vegetation adapts to decreased water supply by adjusting its physiological and ecological strategies, thereby increasing its sensitivity to precipitation (Sun et al. 2024). Although the contributions of these climatic factors fluctuated more in the SH than NH, the average contributions of temperature and radiation in the SH remained greater than in the NH. In addition, the contribution of VGC_EOS to SOS was marginal compared to climatic factors, and it increased across a series of moving window, shifting from advancing to delaying SOS. The opposite trend occurred for shrublands. Once vegetation entered dormancy, the duration of dormancy and prevailing environmental conditions probably exerted a greater influence on the next year's SOS than VGC_EOS. The SOS is determined primarily by climatic conditions in the spring of that year, which affect plant metabolism and growth recovery (Shen et al. 2020). The shift from an advance to a delay of SOS, influenced by the previous year's EOS, may be attributed to global climate warming and alterations in seasonal patterns (Shen et al. 2015). Shrublands may react more quickly to temperature and radiation changes than other vegetation types, resulting in differing SOS responses (Flynn & Wolkovich 2018).

The contribution of precipitation and radiation to delay EOS is increasing in savannas. Savannas are generally located in semi-arid regions, where water is one of the main limiting factors for vegetation growth (Neimane-Sroma et al. 2024). Therefore, an increase in precipitation can delay the EOS. Radiation maintains the hormonal balance in plants, playing a key role in regulating their responses to seasonal changes (Flynn & Wolkovich 2018). On a global scale, VGC_SOS has a substantial impact on EOS. It also increased its contribution to delay EOS over time for both hemispheres and all vegetation types. The contribution of VGC_SOS to EOS exhibited an increasing trend over the two periods, with the largest contribution in forests and the smallest in grasslands. The main reasons for these occurrences are: 1) the advancing rise in spring temperatures can stimulate plants from dormancy and begin photosynthesis and a new growth cycle. Plants that initiate growth earlier might also reach maturity faster, potentially advancing the timing of autumn phenology due to accelerated life cycles. Thus, a warmer climate could uniformly shift the phenological events, maintaining the synchronization of life cycle stages across the growing season (Shen et al. 2015; Meng et al. 2021; Pandiyan et al. 2022). 2) An early start to growth in spring provides plants with more time to accumulate energy and nutrients through photosynthesis (Hurford et al. 2019a). This increased energy reserve enables plants to maintain greater activity levels throughout the growing season (Tang et al. 2024), including during the autumn when they would typically begin to enter dormancy (Liu et al. 2016b; Lian et al. 2021). Therefore, the delayed EOS in autumn can be seen as a direct result of the earlier onset of activity in spring. The length of the growing season in temperate forests has been increasing under recent climate change because of earlier leaf unfold and later leaf senescence (Delpierre et al. 2009). However, Zani et al. (2020) reported that this trend might be reversed as increasing photosynthetic productivity could drive earlier autumn leaf senescence (Zani et al. 2020), a view that is consistent with this study. An increasing number of regions experienced increases in the impact of VGC_SOS on vegetation EOS. This contribution exhibited an increasing trend over time in the NH, but not in the SH. Possible reasons include: firstly, over the past four decades, precipitation in the SH has been gradually decreasing, while evaporation has been increasing, resulting in the region experiencing "warming and drying" climate conditions. In this case, these conditions restrict the growth and physiological activities of vegetation, and weaken the impact of VGC_SOS on vegetation EOS. This phenomenon can be reasonably explained by the insufficient water supply which may cause the plant to enter a dormant period and, consequently, earlier vegetation EOS. Nevertheless, these results further suggest widespread increases in the connection between SOS and EOS, which may have resulted from the intrinsic adjustments of plants to adapt to the changing climates. Furthermore, the fundamental differences in the climatic conditions between the NH and SH are due to geographic distribution (Poulter et al. 2014), land-sea ratio, and heterogeneity in vegetation distribution (Bonachela et al. 2021). These differences, combined with geographic variations in climate change and the ecological responses and adaptability of vegetation, could have caused the temporal differences between the hemispheres (Hurford et al. 2019b).

Other than the influential factors considered above, vegetation growth is also affected by human activities and natural disturbances, such as land-use changes (Sterling et al. 2012), fires (Brando et al. 2014), and plant diseases (Jiang et al. 2017b). Including these factors in the analyses would yield more comprehensive results.

In this study, we used GIMMS NDVI3g (1982–2015) and MODIS NDVI (2001–2022) datasets, and examined global vegetation phenological shifts, emphasizing the relative impacts of climatic factors and VGC. In the NH, SOS was driven mainly by temperature and radiation, whereas EOS was driven mainly by temperature and precipitation, with VGC_SOS as a secondary but notable factor. In the SH, radiation was the key driver of SOS, while VGC_SOS was the primary driver of EOS. The role of VGC_SOS in influencing EOS has increased substantially over the past four decades, particularly in forests, underscoring its growing importance relative to other climatic factors. These findings suggest that VGC_SOS is a critical factor affecting EOS in both the NH and SH. Therefore, integrating VGC_SOS into phenology models can improve the accuracy of predictions related to vegetation phenology dynamics and carbon sequestration under future climate change scenarios.

Satellite-based NDVI datasets

We employed the Global Inventory Modeling and Mapping Studies (GIMMS) NDVI3g dataset to extract vegetation phenology from 1982 to 2015. This dataset is compiled from multi-band reflectance acquired by the Advanced Very High-Resolution Radiometer (AVHRR) sensors onboard the National Oceanic and Atmospheric Administration (NOAA) (Liu et al. 2012). GIMMS NDVI3g data have been applied extensively on global change research encompassing vegetation dynamics, drought monitoring, ecological modeling, and climate change impact assessments (Julien & Sobrino 2009). There period from 1982 to 2015 was covered with updates every 15 days, with a spatial resolution of approximately 8 km (1/12 degree) (https://ecocast.arc.nasa.gov/data/pub/gimms/3g.v0/).

The other data used for phenology were extracted from the Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI dataset. It is a comprehensive collection product derived from the MODIS sensor aboard NASA's Terra satellite (MOD13). Version 6.1, the latest update, provides advancements in data processing and quality assurance protocols. The preprocessing of MODIS NDVI data is intricate, incorporating cloud detection, atmospheric corrections, and rigorous quality checks to ensure the reliability of the time-series data (Fensholt et al. 2009). Given its superior temporal and spatial resolutions, the MODIS NDVI dataset excels in agricultural monitoring, vegetation change tracking, ecosystem dynamics studies, and global change surveillance (Liu et al. 2012). Spanning from 2001 to 2022, this dataset provides data with a spatial resolution of 0.05° and a temporal resolution of 16 days, resulting in 23 observational periods per year (https://lpdaac.usgs.gov/products/mod13c1v006/).

Climate dataset

ERA5, the fifth-generation atmospheric reanalysis dataset by the European Centre for Medium-Range Weather Forecasts (ECMWF), has covered global climate data since January 1950. Developed by ECMWF's Copernicus Climate Change Service (C3S), ERA5 provides hourly updates of an extensive array of climatic factors, including atmospheric, terrestrial, and oceanic components. The dataset spans the Earth with a 30 km resolution grid and extends through 137 vertical levels from the surface to 80 km in height. It also provides uncertainty estimates for all variables at diminished spatial and temporal resolutions (Gill et al. 2015). By integrating model data with worldwide observational data, ERA5 creates a globally cohesive and consistent dataset, superseding its predecessor, ERA-Interim. For this research, we used monthly ERA5 data for 2 m air temperature (Tem) (Gill et al.), precipitation (Pre), surface net solar radiation (Rad), and potential evapotranspiration (PET) from 1982 to 2022, with a spatial resolution of 0.1° (https://confluence.ecmwf.int/display/CKB/ERA5).

Land cover dataset

MCD12C1 is a global land cover classification dataset developed by the Earth Observing System project of NASA. This dataset uses data from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensors aboard the Terra and Aqua satellites, which pass in the morning and afternoon, respectively, offering comprehensive diurnal coverage (Friedl et al. 2010). The MODIS Land Cover Type MCD12C1 dataset is produced through a supervised decision tree classification method, featuring a spatial resolution of 0.05°. It comprises 17 land cover classes (https://lpdaac.usgs.gov/products/mcd12c1v006). MCD12C1 is applied widely in climate change research, ecosystem modeling (Sun et al. 2024), carbon and hydrological cycle studies (Kroner & Way 2016), and evaluation of human activities' impact on the environment (Xiong et al. 2023). In this study, four major vegetation types were considered: forests, shrublands, savannas, and grasslands (Figure S1). The forest types included evergreen needleleaf, evergreen broadleaf, deciduous needleleaf, deciduous broadleaf and mixed forests; and the shrubland types included open and closed shrublands. Croplands were excluded, as their growth is governed predominantly by human management practices (e.g., irrigation, fertilization, cultivation timing, and multiple cropping), rather than by environmental factors (Chen et al. 2018).

Phenology extraction method

We selected two datasets to extract phenology: GIMMS NDVI3g (1982–2015) and MODIS NDVI (2001–2022); NDVI products are influenced by climatic conditions, snow, and cloud cover (Tarnavsky et al. 2008). Accordingly, we adopted a strict processing procedure: initially, the area covered with barren or sparse vegetation was excluded based on the criteria of an annual average NDVI value below 0.1 (Wang et al. 2020). Subsequently, the Savitzky-Golay (S-G) filter smoothed the NDVI time series to mitigate noises from atmospheric conditions, clouds, and snow cover (Wang et al. 2019). Then, we employed the double logistic function (Gonsamo et al. 2012) to extract SOS and EOS:

$${\text{NDVI(t)}}={\alpha _1}+\frac{{{\alpha _2}}}{{1+{e^{ - {\partial _1}(t - {\beta _1})}}}} - \frac{{{\alpha _3}}}{{1+{e^{ - {\partial _2}(t - {\beta _2})}}}}$$

where t is the day of year (DOY), α₁ represents the NDVI background value, α₂ is the difference between the summer NDVI peak and the spring minimum, and α₃ is the difference between the summer NDVI peak and the autumn minimum. ə₁ and ə₂ are the curvature parameters for NDVI increase and decrease, respectively, while β₁ and β₂ represent the mean dates of leaf unfolding in spring and leaf senescence in autumn, respectively. SOS and EOS are determined as the day-of-year (DOY) corresponding to the local maxima and minima of the first derivative of the NDVI curve, respectively.

Data analyses

The temporal trends of SOS, EOS, and climatic variables were assessed using the non-parametric Mann-Kendall test (Lian et al. 2020) which is employed widely across environmental sciences, meteorology, hydrology, and related fields (Li et al. 2017; Chen et al. 2018). To ensure the study's accuracy and robustness, statistical analyses were limited to pixels without missing phenological values within their respective timeframes in both datasets.

Spearman's rank partial correlation analysis tested the impacts of climatic factors on phenology (SOS and EOS). It concurrently controls for other climatic variables to assess the correlations between phenology and various driving factors. Considering the lag effect inherent in vegetation phenology, phenological events are particularly susceptible to prior climatic influences (Wang & Wu 2019). Our research examined climatic factors at different temporal scales (3, 6, and 12 months prior to phenological occurrences) for each pixel, evaluating the pixel-specific phenological responses to these factors. As the climatic variables at the 3-month scale had the largest impacts on phenology (Supplementary Fig. 5), we used the 3-month climatic factors for later analyses.

The method based on partial derivatives quantified the relative importance of climatic factors and VGC on the trend of phenology (Lian et al. 2021). This data analysis technique has been used extensively to decipher the relative contributions of various climatic and non-climatic factors affecting key aspects of ecosystems, and has assessed impacts on hydrology (Meng & Mo 2011), evapotranspiration (Li et al. 2017), vegetation productivity (Yan et al. 2019), and phenology (Yuan et al. 2020). The long-term trend in phenology is influenced by a combination of climatic factors, VGC, and other factors (Piao et al. 2019). In the present study, VGC_EOS represents the VGC effect of EOS from the previous year on SOS in the current year, while VGC_SOS represents the impact of SOS on EOS in the current year.

Thus, taking SOS as an example, the trends of SOS (T_SOS) can be described as:

$$\begin{gathered} {T_{SOS}}=Con(climate)+Con(VG{C_{EOS}})+Con(other) \\ =Co{n_{Tem}}+Co{n_{Pre}}+Co{n_{Rad}}+Co{n_{PET}}+Con(VG{C_{EOS}})+Co{n_{other}} \\ =\frac{{{\delta _{SOS}}}}{{{\delta _{Tem}}}} \times \frac{{{d_{Tem}}}}{{{d_t}}}+\frac{{{\delta _{SOS}}}}{{{\delta _{Pre}}}} \times \frac{{{d_{Pre}}}}{{{d_t}}}+\frac{{{\delta _{SOS}}}}{{{\delta _{Rad}}}} \times \frac{{{d_{Rad}}}}{{{d_t}}}+\frac{{{\delta _{SOS}}}}{{{\delta _{PET}}}} \times \frac{{{d_{PET}}}}{{{d_t}}}+\frac{{{\delta _{SOS}}}}{{{\delta _{EOS}}}} \times \frac{{{d_{EOS}}}}{{{d_t}}}+Co{n_{other}} \\ \end{gathered}$$

where Con(climate) represents the cumulative contribution of temperature (Con_Tem), precipitation (Con_Pre), radiation (Con_Rad), and PET (Con_PET) to SOS, and Con (VGC_EOS) and Con (other) represent the contributions of VGC_EOS and other factors to SOS. Positive values indicate a contribution to the delay of SOS, while negative values indicate a contribution to its advancement. δ_SOS/δ_Tem, δ_SOS/δ_Pre, δ_SOS/δ_Rad, and δ_SOS/δ_PET indicate the sensitivity of SOS to temperature, precipitation, radiation, and potential evapotranspiration, respectively. These results can be obtained from the slopes of multiple linear regression analysis. d_Tem/d_t, d_Pre/d_t, d_Rad/d_t, d_PET/d_t, and d_SOS/d_t represent the interannual changing rates of temperature, precipitation, radiation, potential evapotranspiration, and SOS, respectively, which can be estimated using the Mann-Kendall slope estimator. The trend calculation for EOS is presented in Eq. (2).

Subsequently, the contribution rates of climatic factors and VGC to the trend of phenology can be estimated through their respective contribution ratios. For example, Eq. (3) describes the contribution rate of VGC_SOS to the EOS trend.

$$Rat{e_{(VG{C_{SOS}})}}=\frac{{Con(VG{C_{SOS}})}}{{Con(VG{C_{SOS}})+Co{n_{Tem}}+Co{n_{Pre}}+Co{n_{Rad}}+Co{n_{PET}}}}$$

Where Rate (VGC_SOS) denotes the contribution ratio of VGC_SOS to the EOS trend. Con (VGC_SOS) represents the contribution of VGC_SOS to the EOS trend, same as Eq. (2). Positive values indicate that the contribution of VGC_SOS is consistent with the trend direction of EOS, while negative values represent the opposite direction. A similar method was used to calculate the contribution rate of climatic factors and VGC_EOS to the SOS trend.

We used a moving window model to detect the spatio-temporal variation characteristics of the contributions of major climatic factors (temperature, precipitation, and radiation) and VGC to phenology (SOS and EOS) over the past forty years, as well as the variation patterns across vegetation types in both hemispheres. We selected a window size of 15 years with a time step of 1 year (Wu et al. 2022). Within each fixed window size, we used nonlinear regression models to assess the contributions of major climatic factors (temperature, precipitation, and radiation) and VGC to the temporal changes in phenology (Wu et al. 2022). Taking EOS as an example:

EOS = a Con_Tem + b Con_Pre + c Con_Rad + d Con(VGC_SOS) + ɛ (5)

where EOS is the EOS trend within the fixed moving window; Con_Tem, Con_Pre, and Con_Rad represent contributions of temperature, precipitation, and radiation, respectively, in the three months preceding the occurrence of EOS during the fixed moving window; Con (VGC_SOS) represents the contribution of SOS's VGC to EOS within fixed moving window; a, b, c and d are regression coefficients, and ε denotes the residuals. Before the nonlinear regression analysis, we detrended the phenology and climatic factors extracted from the two datasets, GIMMS (1982–2015) and MODIS (2001–2022).

Accordingly, the temporal variation trend of contribution (Tcon) was derived from the slope of the least squares regression of contributions across a series of moving windows. Significantly increasing and decreasing trends in Tcon are denoted by Tcon + and Tcon − at P < 0.05, respectively, while NS indicates a non-significant trend.

Funding information

The National Key R&D Program of China (No. 2023YFF0805600, 2023YFF0805602, 2023YFF0805604) and the National Natural Science Foundation of China (No. 42201041, 42471046).

Acknowledge

This study is supported by the National Key R&D Program of China (No. 2023YFF0805600, 2023YFF0805602, 2023YFF0805604) and the National Natural Science Foundation of China (No. 42201041, 42471046).

Author contributions

Conflict of interest statement

The authors declare no conflict of interest.

Supporting information

Additional supporting information can be found “Supplementary.docx”.

An, Y., Gao, W., Gao, Z., Liu, C. & Shi, R. (2014). Trend analysis for evaluating the consistency of Terra MODIS and SPOT VGT NDVI time series products in China. Frontiers of Earth Science, 9, 125-136.
Bonachela, J.A., Burrows, M.T., Pinsky, M.L. & Calcagno, V. (2021). Shape of species climate response curves affects community response to climate change. Ecology Letters, 24, 708-718.
Brando, P.M., Balch, J.K., Nepstad, D.C., Morton, D.C., Putz, F.E., Coe, M.T., Silverio, D., Macedo, M.N., Davidson, E.A., Nobrega, C.C., Alencar, A. & Soares-Filho, B.S. (2014). Abrupt increases in Amazonian tree mortality due to drought-fire interactions. Proc Natl Acad Sci U S A, 111, 6347-6352.
Chang, Q., Xiao, X., Jiao, W., Wu, X., Doughty, R., Wang, J., Du, L., Zou, Z. & Qin, Y. (2019). Assessing consistency of spring phenology of snow-covered forests as estimated by vegetation indices, gross primary production, and solar-induced chlorophyll fluorescence. Agricultural and Forest Meteorology, 275, 305-316.
Chen, L., Huang, J.G., Ma, Q., Hänninen, H., Tremblay, F. & Bergeron, Y. (2018). Long‐term changes in the impacts of global warming on leaf phenology of four temperate tree species. Global Change Biology, 25, 997-1004.
Chen, S., Fu, Y.H., Geng, X., Hao, Z., Tang, J., Zhang, X., Xu, Z. & Hao, F. (2022). Influences of Shifted Vegetation Phenology on Runoff Across a Hydroclimatic Gradient. Frontiers in Plant Science, 12.
Chen, Y., Zhang, L., Shi, X., Ban, Y., Liu, H. & Zhang, D. (2019). Life history responses of spring-and autumn-germinated ephemeral plants to increased nitrogen and precipitation in the Gurbantunggut Desert. Science of The Total Environment, 659, 756-763.
De Pue, J., Wieneke, S., Bastos, A., Barrios, J.M., Liu, L., Ciais, P., Arboleda, A., Hamdi, R., Maleki, M., Maignan, F., Gellens-Meulenberghs, F., Janssens, I. & Balzarolo, M. (2023). Temporal variability of observed and simulated gross primary productivity, modulated by vegetation state and hydrometeorological drivers. Biogeosciences, 20, 4795-4818.
Delpierre, N., Dufrêne, E., Soudani, K., Ulrich, E., Cecchini, S., Boé, J. & François, C. (2009). Modelling interannual and spatial variability of leaf senescence for three deciduous tree species in France. Agricultural and Forest Meteorology, 149, 938-948.
Descals, A., Verger, A., Yin, G., Filella, I., Fu, Y.H., Piao, S., Janssens, I.A. & Penuelas, J. (2023). Radiation-constrained boundaries cause nonuniform responses of the carbon uptake phenology to climatic warming in the Northern Hemisphere. Glob Chang Biol, 29, 719-730.
Fensholt, R. & Proud, S.R. (2012). Evaluation of Earth Observation based global long term vegetation trends — Comparing GIMMS and MODIS global NDVI time series. Remote Sensing of Environment, 119, 131-147.
Fensholt, R., Rasmussen, K., Nielsen, T.T. & Mbow, C. (2009). Evaluation of earth observation based long term vegetation trends — Intercomparing NDVI time series trend analysis consistency of Sahel from AVHRR GIMMS, Terra MODIS and SPOT VGT data. Remote Sensing of Environment, 113, 1886-1898.
Flynn, D.F.B. & Wolkovich, E.M. (2018). Temperature and photoperiod drive spring phenology across all species in a temperate forest community. New Phytol, 219, 1353-1362.
Friedl, M.A., Sulla-Menashe, D., Tan, B., Schneider, A., Ramankutty, N., Sibley, A. & Huang, X. (2010). MODIS Collection 5 global land cover: Algorithm refinements and characterization of new datasets. Remote Sensing of Environment, 114, 168-182.
Fu, Y.H., Zhao, H., Piao, S., Peaucelle, M., Peng, S., Zhou, G., Ciais, P., Huang, M., Menzel, A., Penuelas, J., Song, Y., Vitasse, Y., Zeng, Z. & Janssens, I.A. (2015). Declining global warming effects on the phenology of spring leaf unfolding. Nature, 526, 104-107.
Fu, Y.S., Campioli, M., Vitasse, Y., De Boeck, H.J., Van den Berge, J., AbdElgawad, H., Asard, H., Piao, S., Deckmyn, G. & Janssens, I.A. (2014). Variation in leaf flushing date influences autumnal senescence and next year's flushing date in two temperate tree species. Proc Natl Acad Sci U S A, 111, 7355-7360.
Ge, Q., Wang, H., Rutishauser, T. & Dai, J. (2015). Phenological response to climate change in China: a meta-analysis. Glob Chang Biol, 21, 265-274.
Gill, A.L., Gallinat, A.S., Sanders-DeMott, R., Rigden, A.J., Short Gianotti, D.J., Mantooth, J.A. & Templer, P.H. (2015). Changes in autumn senescence in northern hemisphere deciduous trees: a meta-analysis of autumn phenology studies. Ann Bot, 116, 875-888.
Gillespie, L.M. & Volaire, F.A. (2017). Are winter and summer dormancy symmetrical seasonal adaptive strategies? The case of temperate herbaceous perennials. Annals of Botany, 119, 311-323.
Gonsamo, A., Chen, J.M., Price, D.T., Kurz, W.A. & Wu, C. (2012). Land surface phenology from optical satellite measurement and CO2 eddy covariance technique. Journal of Geophysical Research: Biogeosciences, 117.
Guay, K.C., Beck, P.S., Berner, L.T., Goetz, S.J., Baccini, A. & Buermann, W. (2014). Vegetation productivity patterns at high northern latitudes: a multi-sensor satellite data assessment. Glob Chang Biol, 20, 3147-3158.
Guo, J., Ma, Q., Xu, H., Luo, Y., He, D., Wang, F., Wu, J., Fu, Y.H., Liu, J., Zhang, R. & Chen, L. (2023). Meta-analytic and experimental evidence that warmer climate leads to shift from advanced to delayed spring phenology. Agricultural and Forest Meteorology, 342.
Hong, S., Zhang, Y., Yao, Y., Meng, F., Zhao, Q. & Zhang, Y. (2022). Contrasting temperature effects on the velocity of early- versus late-stage vegetation green-up in the Northern Hemisphere. Glob Chang Biol, 28, 6961-6972.
Huang, Z., Zhou, L. & Chi, Y. (2023). Spring phenology rather than climate dominates the trends in peak of growing season in the Northern Hemisphere. Global Change Biology, 29, 4543-4555.
Hurford, A., Cobbold, C.A. & Molnar, P.K. (2019a). Skewed temperature dependence affects range and abundance in a warming world. Proc Biol Sci, 286, 20191157.
Hurford, A., Cobbold, C.A. & Molnár, P.K. (2019b). Skewed temperature dependence affects range and abundance in a warming world. Proceedings of the Royal Society B: Biological Sciences, 286.
Jiang, C., Ryu, Y., Fang, H., Myneni, R., Claverie, M. & Zhu, Z. (2017a). Inconsistencies of interannual variability and trends in long-term satellite leaf area index products. Glob Chang Biol, 23, 4133-4146.
Jiang, L., Guli, J., Bao, A., Guo, H. & Ndayisaba, F. (2017b). Vegetation dynamics and responses to climate change and human activities in Central Asia. Sci Total Environ, 599-600, 967-980.
Julien, Y. & Sobrino, J.A. (2009). Global land surface phenology trends from GIMMS database. International Journal of Remote Sensing, 30, 3495-3513.
Kroner, Y. & Way, D.A. (2016). Carbon fluxes acclimate more strongly to elevated growth temperatures than to elevated CO2 concentrations in a northern conifer. Global Change Biology, 22, 2913-2928.
Li, G., Zhang, F., Jing, Y., Liu, Y. & Sun, G. (2017). Response of evapotranspiration to changes in land use and land cover and climate in China during 2001–2013. Science of The Total Environment, 596-597, 256-265.
Li, P., Peng, C., Wang, M., Luo, Y., Li, M., Zhang, K., Zhang, D. & Zhu, Q. (2018). Dynamics of vegetation autumn phenology and its response to multiple environmental factors from 1982 to 2012 on Qinghai-Tibetan Plateau in China. Science of The Total Environment, 637-638, 855-864.
Li, Z., Li, X. & Yang, S. (2023). Plant drought adaptation strategies regulate alpine grassland water yield in the Qinghai Lake Basin, northeastern Qinghai-Tibet Plateau. Journal of Hydrology: Regional Studies, 48.
Lian, X., Piao, S., Chen, A., Wang, K., Li, X., Buermann, W., Huntingford, C., Penuelas, J., Xu, H. & Myneni, R.B. (2021). Seasonal biological carryover dominates northern vegetation growth. Nat Commun, 12, 983.
Lian, X., Piao, S., Laurent, Z., Yue, L., Chris, H., Philippe, C. & Alessandro, C. (2020). Summer soil drying exacerbated by earlier spring greening of northern vegetation. Sci. Adv., 6, eaax0255.
Liu, Q., Fu, Y.H., Liu, Y., Janssens, I.A. & Piao, S. (2018). Simulating the onset of spring vegetation growth across the Northern Hemisphere. Glob Chang Biol, 24, 1342-1356.
Liu, Q., Fu, Y.H., Zeng, Z., Huang, M., Li, X. & Piao, S. (2016a). Temperature, precipitation, and insolation effects on autumn vegetation phenology in temperate China. Glob Chang Biol, 22, 644-655.
Liu, Q., Fu, Y.H., Zhu, Z., Liu, Y., Liu, Z., Huang, M., Janssens, I.A. & Piao, S. (2016b). Delayed autumn phenology in the Northern Hemisphere is related to change in both climate and spring phenology. Glob Chang Biol, 22, 3702-3711.
Liu, Y., Liu, R. & Chen, J.M. (2012). Retrospective retrieval of long‐term consistent global leaf area index (1981–2011) from combined AVHRR and MODIS data. Journal of Geophysical Research: Biogeosciences, 117.
Liu, Y., Shen, X., Zhang, J., Wang, Y., Wu, L., Ma, R., Lu, X. & Jiang, M. (2023). Spatiotemporal variation in vegetation phenology and its response to climate change in marshes of Sanjiang Plain, China. Ecology and Evolution, 13.
Mei, L., Bao, G., Tong, S., Yin, S., Bao, Y., Jiang, K., Hong, Y., Tuya, A. & Huang, X. (2021). Elevation-dependent response of spring phenology to climate and its legacy effect on vegetation growth in the mountains of northwest Mongolia. Ecological Indicators, 126.
Meng, D. & Mo, X. (2011). Assessing the effect of climate change on mean annual runoff in the Songhua River basin, China. Hydrological Processes, 26, 1050-1061.
Meng, L., Zhou, Y., Gu, L., Richardson, A.D., Peñuelas, J., Fu, Y., Wang, Y., Asrar, G.R., De Boeck, H.J., Mao, J., Zhang, Y. & Wang, Z. (2021). Photoperiod decelerates the advance of spring phenology of six deciduous tree species under climate warming. Global Change Biology, 27, 2914-2927.
Nagai, S., Saitoh, T.M. & Morimoto, H. (2020). Does global warming decrease the correlation between cherry blossom flowering date and latitude in Japan? International Journal of Biometeorology, 64, 2205-2210.
Neimane-Sroma, S., Durand, M., Lintunen, A., Aalto, J. & Robson, T.M. (2024). Shedding light on the increased carbon uptake by a boreal forest under diffuse solar radiation across multiple scales. Glob Chang Biol, 30, e17275.
Nilsson, O. (2022). Winter dormancy in trees. Current Biology, 32, R630-R634.
Pandiyan, S., Govindjee, G., Meenatchi, S., Prasanna, S., Gunasekaran, G. & Guo, Y. (2022). Evaluating the Impact of Summer Drought on Vegetation Growth Using Space-Based Solar-Induced Chlorophyll Fluorescence Across Extensive Spatial Measures. Big Data, 10, 230-245.
Peng, J., Wu, C., Wang, X. & Lu, L. (2021). Spring phenology outweighed climate change in determining autumn phenology on the Tibetan Plateau. International Journal of Climatology, 41, 3725-3742.
Peng, J., Wu, C., Zhang, X., Wang, X. & Gonsamo, A. (2019). Satellite detection of cumulative and lagged effects of drought on autumn leaf senescence over the Northern Hemisphere. Glob Chang Biol, 25, 2174-2188.
Piao, S., Liu, Q., Chen, A., Janssens, I.A., Fu, Y., Dai, J., Liu, L., Lian, X., Shen, M. & Zhu, X. (2019). Plant phenology and global climate change: Current progresses and challenges. Glob Chang Biol, 25, 1922-1940.
Poulter, B., Frank, D., Ciais, P., Myneni, R.B., Andela, N., Bi, J., Broquet, G., Canadell, J.G., Chevallier, F., Liu, Y.Y., Running, S.W., Sitch, S. & van der Werf, G.R. (2014). Contribution of semi-arid ecosystems to interannual variability of the global carbon cycle. Nature, 509, 600-603.
Rossi, S. & Isabel, N. (2017). Bud break responds more strongly to daytime than night-time temperature under asymmetric experimental warming. Glob Chang Biol, 23, 446-454.
Schwartz, M.D., Ahas, R. & Aasa, A. (2006). Onset of spring starting earlier across the Northern Hemisphere. Global Change Biology, 12, 343-351.
Shen, M., Jiang, N., Peng, D., Rao, Y., Huang, Y., Fu, Y.H., Yang, W., Zhu, X., Cao, R., Chen, X., Chen, J., Miao, C., Wu, C., Wang, T., Liang, E. & Tang, Y. (2020). Can changes in autumn phenology facilitate earlier green-up date of northern vegetation? Agricultural and Forest Meteorology, 291.
Shen, M., Piao, S., Cong, N., Zhang, G. & Jassens, I.A. (2015). Precipitation impacts on vegetation spring phenology on the Tibetan Plateau. Global Change Biology, 21, 3647-3656.
Sterling, S.M., Ducharne, A. & Polcher, J. (2012). The impact of global land-cover change on the terrestrial water cycle. Nature Climate Change, 3, 385-390.
Sun, W., Li, J., Yu, R., Li, N. & Zhang, Y. (2024). Exploring changes of precipitation extremes under climate change through global variable-resolution modeling. Sci Bull (Beijing), 69, 237-247.
Tagesson, T., Fensholt, R., Guiro, I., Rasmussen, M.O., Huber, S., Mbow, C., Garcia, M., Horion, S., Sandholt, I., Holm‐Rasmussen, B., Göttsche, F.M., Ridler, M.E., Olén, N., Lundegard Olsen, J., Ehammer, A., Madsen, M., Olesen, F.S. & Ardö, J. (2014). Ecosystem properties of semiarid savanna grassland in West Africa and its relationship with environmental variability. Global Change Biology, 21, 250-264.
Tang, J., Niu, B., Hu, Z. & Zhang, X. (2024). Increasing susceptibility and shortening response time of vegetation productivity to drought from 2001 to 2021. Agricultural and Forest Meteorology, 352.
Tarnavsky, E., Garrigues, S. & Brown, M.E. (2008). Multiscale geostatistical analysis of AVHRR, SPOT-VGT, and MODIS global NDVI products. Remote Sensing of Environment, 112, 535-549.
Vitasse, Y., Signarbieux, C. & Fu, Y.H. (2018). Global warming leads to more uniform spring phenology across elevations. Proc Natl Acad Sci U S A, 115, 1004-1008.
Wang, C., Yang, Y., Yin, G., Xie, Q., Xu, B., Verger, A., Descals, A., Filella, I. & Peñuelas, J. (2024). Divergence in Autumn Phenology Extracted From Different Satellite Proxies Reveals the Timetable of Leaf Senescence Over Deciduous Forests. Geophysical Research Letters, 51.
Wang, X., Dannenberg, M.P., Yan, D., Jones, M.O., Kimball, J.S., Moore, D.J.P., van Leeuwen, W.J.D., Didan, K. & Smith, W.K. (2020). Globally Consistent Patterns of Asynchrony in Vegetation Phenology Derived From Optical, Microwave, and Fluorescence Satellite Data. Journal of Geophysical Research: Biogeosciences, 125.
Wang, X. & Wu, C. (2019). Estimating the peak of growing season (POS) of China’s terrestrial ecosystems. Agricultural and Forest Meteorology, 278.
Wang, X., Xiao, J., Li, X., Cheng, G., Ma, M., Zhu, G., Altaf Arain, M., Andrew Black, T. & Jassal, R.S. (2019). No trends in spring and autumn phenology during the global warming hiatus. Nature Communications, 10.
White, M.A., De Beurs, K.M., Didan, K., Inouye, D.W., Richardson, A.D., Jensen, O.P., O'Keefe, J., Zhang, G., Nemani, R.R., Van Leeuwen, W.J.D., Brown, J.F., De Wit, A., Schaepman, M., Lin, X., Dettinger, M., Bailey, A.S., Kimball, J., Schwartz, M.D., Baldocchi, D.D., Lee, J.T. & Lauenroth, W.K. (2009). Intercomparison, interpretation, and assessment of spring phenology in North America estimated from remote sensing for 1982–2006. Global Change Biology, 15, 2335-2359.
Wu, C., Peng, J., Ciais, P., Peñuelas, J., Wang, H., Beguería, S., Andrew Black, T., Jassal, R.S., Zhang, X., Yuan, W., Liang, E., Wang, X., Hua, H., Liu, R., Ju, W., Fu, Y.H. & Ge, Q. (2022). Increased drought effects on the phenology of autumn leaf senescence. Nature Climate Change, 12, 943-949.
Xie, Y., Wang, X. & Silander, J.A., Jr. (2015). Deciduous forest responses to temperature, precipitation, and drought imply complex climate change impacts. Proc Natl Acad Sci U S A, 112, 13585-13590.
Xiong, Y., Mo, S., Wu, H., Qu, X., Liu, Y. & Zhou, L. (2023). Influence of human activities and climate change on wetland landscape pattern—A review. Science of The Total Environment, 879.
Yan, Y., Liu, X., Wen, Y. & Ou, J. (2019). Quantitative analysis of the contributions of climatic and human factors to grassland productivity in northern China. Ecological Indicators, 103, 542-553.
Ye, W., van Dijk, A.I.J.M., Huete, A. & Yebra, M. (2021). Global trends in vegetation seasonality in the GIMMS NDVI3g and their robustness. International Journal of Applied Earth Observation and Geoinformation, 94.
Yuan, M., Zhao, L., Lin, A., Li, Q., She, D. & Qu, S. (2020). How do climatic and non-climatic factors contribute to the dynamics of vegetation autumn phenology in the Yellow River Basin, China? Ecological Indicators, 112.
Yuan, Y., Bao, A., Jiapaer, G., Jiang, L. & De Maeyer, P. (2022). Phenology-based seasonal terrestrial vegetation growth response to climate variability with consideration of cumulative effect and biological carryover. Sci Total Environ, 817, 152805.
Zani, D., Crowther, T.W., Mo, L., Renner, S.S. & Zohner, C.M. (2020). Increased growing-season productivity drives earlier autumn leaf senescence in temperate trees. Science, 370, 1066-1071.
Zeng, Z., Wu, W., Ge, Q., Li, Z., Wang, X., Zhou, Y., Zhang, Z., Li, Y., Huang, H., Liu, G. & Peñuelas, J. (2021). Legacy effects of spring phenology on vegetation growth under preseason meteorological drought in the Northern Hemisphere. Agricultural and Forest Meteorology, 310.
Zhao, Q., Zhu, Z., Zeng, H., Myneni, R.B., Zhang, Y., Peñuelas, J. & Piao, S. (2022). Seasonal peak photosynthesis is hindered by late canopy development in northern ecosystems. Nature Plants, 8, 1484-1492.
Zhou, X., Xin, Q., Dai, Y., Li, W. & Qiao, H. (2021). A deep‐learning‐based experiment for benchmarking the performance of global terrestrial vegetation phenology models. Global Ecology and Biogeography, 30, 2178-2199.
Zhou, X., Xin, Q., Zhang, S., Delzon, S. & Dai, Y. (2023). A prognostic vegetation phenology model to predict seasonal maximum and time series of global leaf area index using climate variables. Agricultural and Forest Meteorology, 342.

There is NO Competing Interest.

Supplementaryrevied.docx
rs.pdf
Reporting Summary

Download PDF

Version 1

posted

You are reading this latest preprint version

The increased effect of spring leaf unfolding on autumn senescence in the northern and southern hemispheres

Status:

Version 1

Abstract

Figures

Introduction

Results

Spatio-temporal patterns in phenology from GIMMS and MODIS datasets

The relationships between phenology and climatic factors

The contribution of climatic factors and VGC to phenology

The spatio-temporal variation characteristics of the contribution of major climatic factors and VGC

Discussion

Phenological data extracted from two datasets

The dominant roles of climatic factors and VGC in phenology

Spatio-temporal trend of the contributions of dominant factors to phenology

Conclusions

Methods

Satellite-based NDVI datasets

Climate dataset

Land cover dataset

Phenology extraction method

Data analyses

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1