Large-scale Climate Drivers Research Articles

Seasonal WaveNet-LSTM: A Deep Learning Framework for Precipitation Forecasting with Integrated Large Scale Climate Drivers

Seasonal WaveNet-LSTM: A Deep Learning Framework for Precipitation Forecasting with Integrated Large Scale Climate Drivers

Using the Observed Variations of the Start Date of the Rainy Season over Central America for Its Reliable Seasonal Outlook

Abstract We introduce a simple method to define the start and the end of the rainiest part of the year as the first and the last day of the year when the daily rain rate is more or less than the annual mean climatological rain rate for a region or at a given grid point of the rainfall analysis, respectively. A novelty of this work is the adoption of a perturbation technique to generate a total of 1001 ensemble members to account for observational and analysis uncertainties. This allows for a probabilistic estimate of the start and retreat dates of the rainy season at the granularity of the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG), version 6, rainfall analysis over Central America. The seasonal cycle of the IMERG rainfall analysis is also found to verify with in situ observations in the region. Many large-scale climate drivers affect regional rainfall, often with complex interactions that affect the onset date, retreat date, and magnitude of the seasonal rainfall cycle, making it difficult to predict the length or total quantity of seasonal rainfall using climate drivers alone. Once an onset date is established, however, this metric alone can be more indicative of both the length and total seasonal rainfall anomaly than predicting how the climate drivers will interact to affect the quantity and duration of upcoming seasonal rainfall. The local relationships of the start date with seasonal length and rainfall anomaly are leveraged to produce effective seasonal outlooks of the rainy season for the region by just monitoring the start date variations. Significance Statement The start date and retreat date of the rainy season in Central America are defined every year, precisely to a specific date from our proposed definition of it. This is possible because the region exhibits a strong seasonality of the rainfall. As a result, the year-to-year (interannual) variation of the seasonal rainfall during the rainy season is also determined by the variations in the length of the season that are usually overlooked in fixed calendar seasons. We define the start or retreat date of the rainy season as the first or the last day of the year when the daily rain rate exceeds or falls below the average daily rainfall from 2001 to 2022, respectively. We also find that both start and retreat date variations independently influence the length and total seasonal rainfall of the rainy season. Consequently, these relationships are leveraged to provide an outlook of the forthcoming rainy season from the diagnosis of the variations of its onset date, which is shown to be an effective predictor with significant useful seasonal prediction skills.

Prediction of streamflow in Chalakudy River Basin, Kerala, by integrating teleconnection patterns of large-scale atmospheric circulations

ABSTRACT In this study, the relationship between large-scale climatic drivers and streamflow of the Chalakudy River Basin, Kerala, was analysed using methods such as bivariate wavelet coherence (BWC), multiple wavelet coherence (MWC), and partial wavelet coherence (PWC) analysis. The four prominent global climate indices chosen are the Indian Ocean Dipole (IOD), North Atlantic Oscillation (NAO), El Niño Southern Oscillation (ENSO), and Pacific Decadal Oscillation (PDO), along with other local climate drivers. The BWC analysis showed that streamflow and rainfall had a very strong in-phase relationship, whereas the maximum temperature and average temperature showed an anti-phase relationship. In the case of global climate drivers, ENSO has a significant impact on the streamflow of the Chalakudy River Basin. The average wavelet coherence (AWC), which quantifies the teleconnections, confirms the observation with a high coherency of 0.75 between streamflow and rainfall and 0.51 for streamflow and ENSO. Streamflow prediction models were developed using random forest (RF) and artificial neural network (ANN) techniques by considering the influence of significant global and local climatic drivers. It was observed that the RF model performed slightly better than the ANN model, with R = 0.875, NSE = 0.766, RMSE = 23.468, and RSR = 0.524.

Variability in flood frequency in sub-Saharan Africa: The role of large-scale climate modes of variability and their future impacts

Sub-Saharan Africa (SSA) is strongly affected by flood hazards, endangering human lives and economic stability. However, the role of internal climate modes of variability in driving fluctuations in SSA flood occurrence remains poorly documented and understood. To address this gap, we quantify the relative and combined contribution of large-scale climate drivers to seasonal and regional flood occurrence using a new 65-year daily streamflow dataset, sea-surface temperatures derived from observations, and 12 Single Model Initial-condition Large Ensembles (SMILEs) from the Coupled Model Intercomparison Project Phases 5 and 6. We find significant relationships between floods and large-scale climate variability across SSA, with climatic drivers accounting for 30–90 % of the variability in floods. Notably, western, central, and the summer-rain region of southern Africa display stronger teleconnections to large-scale climate variability in comparison to East Africa and the winter-rain region of South Africa, where regional circulation patterns and human activities may play a more important role. In southern and eastern Africa, floods are mainly influenced by teleconnections with the Pacific and Indian Oceans, while in western and central Africa, teleconnections with the Atlantic Ocean and Mediterranean Sea play a larger role. We also find that the number of floods is projected to fluctuate by ± 10–50 % during the 21st century in response to different sequences of key modes of climate variability. We also note that the relative contributions of large-scale climate variability to future flood risks are generally consistent across all SMILEs. Our findings thus provide valuable information for long-term disaster risk reduction and management.

Influence of large scale climate drivers on hydro-climate variability in Dwangwa River Basin, Malawi

Large-scale climate processes such as the Indian Ocean Dipole (IOD) have significant roles in modulating rainfall and hydrological systems. Understanding such processes can inform adaptation measures for climate change and variability, as well as water resource management and planning. This study investigated the impact of the Indian Ocean Dipole (IOD) on rainfall and discharge variability in the Dwangwa River Basin (DRB) in Malawi, a key inflow to Lake Malawi. Specifically, the study analysed annual rainfall variability trends from 1985 to 2015 using the Coefficient of Variation (CV) and the annual Precipitation Concentration Index (PCI). The significance and direction of rainfall and discharge trends were quantified using the Mann–Kendall trend test at α = 0.05 significance level. To evaluate the association between rainfall and IOD, the Pearson product moment used three IOD phases: positive, negative, and neutral. Simple linear regression was utilised to check the response of the river during the concerned IOD phases. The study found CVs between 20 and 30%, typical of climates with moderate monthly rainfall variability. The PCI ranged from 20 to 30%, suggesting a strongly seasonal and highly variable temporal intra-annual rainfall distribution in the DRB. Moreover, the Mann–Kendall test statistics showed insignificant decrease in annual rainfall trends. Further, the findings demonstrated an insignificant negative correlation between rainfall and positive IOD, with rainfall increases associated with negative IOD, whereas positive IOD is associated with decreased river discharge. Consequently, El Niño and a positive IOD could cause DRB to have low water availability. Therefore, the study demonstrates that rainfall is experiencing a decreasing trend, which is driven by large-scale mechanics.

Influence of global teleconnections on long-term variability in flood seasonality across peninsular India

Climate change and natural climate variability impact not only the frequency and magnitude of floods but also flood seasonality. However, limited to no study has investigated the seasonality in flood occurrence in peninsular Indian rivers. The Central Water Commission opening its long-term records of river stage and discharge gathered from many monitoring stations offers a unique opportunity to analyse flood seasonality. The primary aims of this study are to identify the time of the year when floods are most likely, investigate whether the occurrence of floods has changed over time due to the influence of climate change and natural climate variability, and determine the most significant large-scale and local climate drivers of flood seasonality. Stage and discharge data from 281 gauging stations across ten peninsular river basins are considered to identify the annual maximum gauge level for a gauging point while retaining the corresponding Julian day (also known as the date of occurrence or DO) for further analysis. Several attributes related to circular statistics are calculated from the DO series to find that 68% of the gauging stations experience floods during August, indicating the dominance of the monsoon system on DO. Preliminary analysis confirms that while most stations experience a non-stationarity in the DO series, a few stations exhibit a non-linear trend. Furthermore, our study develops a machine learning model with DO as predictand and 36 local and large-scale predictors to identify the dominant predictors of DO. The predictor importance metric shows that the Pacific Decadal Oscillation (PDO) and the El Niño and Southern Oscillation (ENSO) explain DO variability better than other drivers. Moreover, since ENSO and PDO are teleconnected with the onset of monsoon and annual maximum rainfall (Pradhan et al 2017 Sci. Rep. 7 14229; Choudhury et al 2021a Q. J. R. Meteorol. Soc. 147 3111–23), flood seasonality demonstrates a connection with both PDO and ENSO.

Investigation of the Historical Trends and Variability of Rainfall Patterns during the March–May Season in Rwanda

This study explores the spatiotemporal variability and determinants of rainfall patterns during the March to May (MAM) season in Rwanda, incorporating an analysis of teleconnections with oceanic–atmospheric indices over the period 1983–2021. Utilizing the Climate Hazards Group Infrared Precipitation with Stations (CHIRPS) dataset, the study employs a set of statistical tools including standardized anomalies, empirical orthogonal functions (EOF), Pearson correlation, the Mann–Kendall (MK) trend test, and Sen’s slope estimator to dissect the intricacies of rainfall variability, trends, and their association with large-scale climatic drivers. The findings reveal a distinct southwest to northwest rainfall gradient across Rwanda, with the MK test signaling a decline in annual precipitation, particularly in the southwest. The analysis for the MAM season reveals a general downtrend in rainfall, attributed in part to teleconnections with the Indian Ocean Sea surface temperatures (SSTs). Notably, the leading EOF mode for MAM rainfall demonstrates a unimodal pattern, explaining a significant 51.19% of total variance, and underscoring the pivotal role of atmospheric dynamics and moisture conveyance in shaping seasonal rainfall. The spatial correlation analysis suggests a modest linkage between MAM rainfall and the Indian Ocean Dipole, indicating that negative (positive) phases are likely to result in anomalously wet (dry) conditions in Rwanda. This comprehensive assessment highlights the intricate interplay between local rainfall patterns and global climatic phenomena, offering valuable insights into the meteorological underpinnings of rainfall variability during Rwanda’s critical MAM season.

Modelling non-stationarity in extreme rainfall using large-scale climate drivers

Flood estimates used in engineering design are commonly based on intensity–duration–frequency (IDF) curves derived from historical extreme rainfalls. Under global warming these extreme rainfalls are increasing, threatening the capacity of existing infrastructure to resist failure as IDF curves traditionally assume no change in rainfall magnitude. Hence, there is a need to investigate the implications of non-stationarity in extreme rainfalls used to derive IDFs across storm durations and annual exceedance probabilities (AEPs). One way of doing this is to incorporate covariates into the fitted probability distribution. However, there are few studies which examine non-stationarity in extreme rainfall using large-scale climate drivers as covariates, with little consensus on which covariate is the most appropriate. Here we evaluate non-stationarity in extreme rainfall across different durations, from the 1 in 5 AEP to the 1 in 100 AEP, using potential large-scale climate drivers including global and continental mean temperature, continental diurnal temperature range, continental dewpoint temperature, continental precipitable water, the Indian Ocean Dipole, the El Niño Southern Oscillation, and the Southern Annular Mode. These covariates are linked to the three parameters of the Generalized Extreme Value distribution to identify the most appropriate form of non-stationary probability model. We analyse 16 different storm durations from 6 min to 7 day annual maxima across 46 stations in Australia. Based on the Akaike Information Criteria, precipitable water is the superior covariate at a large proportion of stations irrespective of storm duration. However, when the modelled quantile changes are inspected, only global temperature is able to adequately capture the variability in changes across both storm duration and annual exceedance probability. Further, when regional average values of mean temperature, diurnal temperature range, dewpoint temperature and precipitable water were used as covariates, there was no improvement in the model performance compared to continental-wide covariates, particularly for short durations. The results using a mean global temperature covariate show that stationary historical rainfall quantile estimates are underestimated by 12 % – 9 % for frequent (1 in 5 AEP) and 23 % – 13 % for rare (1 in 100 AEP) short duration (6 min – 30 min) events. Moving forward our results suggests infrastructure design needs to incorporate non-stationarity in IDFs to ensure infrastructure and flood planning levels are not under-designed.

Examining the effects of climate change and human impacts on a high-resolution, late Holocene paleofire record from South Africa's winter rainfall zone

Fire is central to the Cape Floristic Region's highly biodiverse and disturbance-adapted Fynbos Biome. However, prehistoric fire regimes, their ecological consequences, and their relationships with large-scale climate drivers and human activities remain poorly understood. Here, we use a high-resolution sedimentary charcoal record from Verlorenvlei, a coastal lake situated on the west coast, to interrogate links between fire, climate, and pastoralism in the Fynbos Biome. Our record has a robust chronology supported by 24 radiocarbon dates and provides a continuous sedimentary sequence spanning the last 4200 years, documenting fire activity before and after the local arrival of pastoralists in the Verlorenvlei area ∼1500 cal years BP. Fire at Verlorenvlei over the last 4200 years is variable, with relatively low activity until ∼2000 cal years BP, after which variable but generally higher fire activity occurs until the highest period of fire activity from ∼1450 to 1800 CE (∼500–150 cal years BP). The increase in fire activity ∼2000 years ago corresponds with a shift in the diatom assemblage at Verlorenvlei from marine towards brackish and freshwater species, reflecting increased precipitation derived from a strengthening of the southern westerly winds. The peak in fire activity beginning ∼1450 CE (∼500 cal years BP), near the onset of the Little Ice Age, tracks a second diatom-inferred strengthening of the westerly winds. Other southern hemisphere and Antarctic records further corroborate this increased westerly influence after ∼2000 years. Linear regression modeling on the fire record indicates that moisture availability is the primary driver of fire at Verlorenvlei, with little evidence that human populations influenced fire. Our reconstruction suggests that fire activity at Verlorenvlei is limited by moisture availability and that wetter conditions facilitate increased vegetation (i.e., fuel) and intensified fire at this otherwise fuel-limited site. This work has implications for management and conservation decisions in response to future predictions of a warmer and drier climate along South Africa's west coast.

Large-scale climatic drivers for warm-season compound drought and heatwave frequency over North China

Based on multiple observational and reanalysis datasets, this study investigated the large-scale climatic drivers responsible for the warm season (April to September) compound drought and heatwave (CDHW) frequency over the North China (NC) domain, both locally and remotely. The warm-season NC CDHW frequency exhibited salient year-to-year variation. We identified that the extensive equivalent barotropic high-pressure anomaly centered around Lake Baikal could be the predominant local-scale driving factor. In addition, the low-pressure circulation anomaly to the southeast played a secondary dynamic amplification role. Such a dipole pattern can jointly cause more frequent occurrences of NC CDHWs by triggering anomalies in localized meteorological variables (e.g., surface warming and drying) and land surface variable (reduced soil moisture) that are favorable for CDHW formation during the warm season. Regarding remote driving factors, central eastern Maritime Continent (MC) convection centered over New Guinea could be an influential and direct atmospheric forcing factor. Meanwhile, the sea surface temperature (SST) anomaly pattern over the tropical eastern Pacific (TEP) was identified as a crucial oceanic forcing indirectly modulating the NC CDHW frequency through the critical medium of MC convection. Through a Hadley-like meridional overturning circulation pattern, the suppressed MC convection can remotely induce a dipolar pattern similar to that related to a higher NC CDHW frequency, exerting a so-called direct modulation effect. Furthermore, positive TEP SST anomalies act as anomalous Rossby wave sources via anomalous upper-tropospheric divergence, characterized by westward-propagating descending Rossby waves with notable wave energy advecting westward toward the central–eastern MC. In such a scenario, a reversed Walker-like circulation pattern can form to sustain the suppressed MC convection, playing an intermediate role in the maintenance of the high-pressure anomaly centered around Lake Baikal. Our findings may provide new insights into the synergistic roles of distant climatic driving factors in the modulation and maintenance of interannual variations of warm-season NC CDHW.

Controls on the stable isotope composition of daily precipitation in Sydney Australia: 9 years of daily data, including Radon-222

Daily and weekly rainfall stable isotopic data, collected over a 9-year period (for daily) and 14-year period (for weekly) at Lucas Heights, Australia, were compared. Small differences in the local meteoric water lines (LMWLs) were seen when developed using daily, weekly, or monthly data (where the effect of short-term small atmospheric variability is reduced by the associated averaging). Although, if a precipitation weighted least squares method was used to derive the LMWLs, the differences between the daily and weekly derived LMWLs were lower.A systematic analysis was then conducted of the impact on the isotopic composition of precipitation from environmental conditions at the sampling site and along the air mass trajectory, the synoptic weather systems, and finally large-scale climate drivers (e.g.,El Niño-Southern Oscillation).The best correlation of δ18O was with local temperature, rainout of the air mass, and rainfall amount at the site. The correlation between precipitation amount and δ18O values was higher when daily samples were used, than when the precipitation weighted monthly values were considered (-0.36 as opposed to −0.31). A good correlation was also seen between δ18O and duration of the precipitation event (r = -0.42). The correlation between δ18O and the fraction of convective rainfall was significant, but low (r = 0.11).On a synoptic systems basis, the lowest δ values were seen in rainfall from offshore low-pressure systems, located to the east or north-east of the site; although, the largest rainfall was seen from low pressure systems arriving from the north-east. The influence of the large-scale climate drivers on the isotopic composition of precipitation was insignificant at this coastal site.When Radon-222 was used to categorise the atmospheric mixing state, rainfall amount was clearly negatively correlated with nocturnal stability, with the lowest amount of rainfall occurring on days with the most stable nocturnal periods. Most of the convective rainfall occurred on these same days, occurring after the warmer period of the day, although the total rainfall amount was low. Analysing the data within the separate radon-derived stability categories improved the correlations with rainfall amount for some of the stability categories.

Impacts of large scale climate drivers on precipitation in Sindh, Pakistan using machine learning techniques

Sindh province of Pakistan has a long history of severe droughts. Several large scale climate drivers (LSCD) are known for their effect on precipitation worldwide but studies in the Sindh region are missing; wide variety of LSCDs and lagged associative information. This study aimed to identify the significant LSCDs in Sindh province of Pakistan and improve the forecast skill of monthly precipitation by employing the principal component analysis (PCA), artificial neural network (ANN), Bayesian regularization neural network (BRNN), and multiple regression analysis (MRA), while considering the 12 months lagged LSCDs such as Nino-1+2, Nino-3, Nino-3.4, Nino-4, Quasi-Biennial Oscillation (QBO) at 30 and 50hPa (QBOI and QBOII), sea surface temperature (SST), 2m air temperature (T2M), 500 hPa and 850 hPa geopotential heights (H500 and H850), surface and 500 hPa zonal velocity (SU and U500), latent and sensible heat fluxes over land (LHFOL and SHFOL), and surface specific humidity (SSH). Global Land Data Assimilation System (GLDAS), Tropical Rainfall Measuring Mission (TRMM), ModernEra Retrospective Analysis for Research and Application (MERRA-2), NOAA, Freie University Berlin, and Hadley Centre Sea Ice and Sea Surface Temperature (HadISST) datasets were used. Results manifested that significant LSCDs with 99% confidence level were SU, U500, T2M, SST, SHFOL, LHFOL, SSH, and H850. During test period, compared with MR models of 0.39 to 0.64 and principal components of 0.31 to 0.57, the ANN and BRNN models had better predictive skills with correlation coefficients of 0.57 to 0.83 and 0.52 to 0.76, respectively. It can be concluded that the ANN and BRNN models enable us to predict monthly precipitation in Sindh region with lagged LSCDs.

Synergistic effects of climate and salinity on radial growth of Excoecaria agallocha L. in the Sundarbans world heritage mangrove ecosystem

The Sundarbans World Heritage ecosystem provides valuable ecosystem goods and services to >7.5 million coastal people in South Asia. This sentinel mangrove ecosystem is threatened by historical anthropogenic disturbances, salinity intrusion and long-term environmental changes including sea level rise. Understanding which factors regulate mangroves’ growth is crucial for taking effective management decisions under global changes. However, a paucity of long-term mangrove tree growth data has limited our ability to quantify and predict the impacts of global changes on mangrove ecosystems. Using long-term growth data archived from the stem discs, this study, therefore, aims to understand how climatic and habitat variables influence radial growth dynamics in Excoecaria agallocha, a hyperdominant mangrove species in the Bangladesh Sundarbans. We find E. agallocha forms a distinct tree-ring boundary characterized by a narrow (2–4 cells wide) band of radially flattened fibres. The crossdatable tree-rings enabled us to develop two ecological zone-specific (low and high saline zone) chronologies (1972–2020) covering the entire salinity stress gradient present in the Sundarbans. Our analyses show: (I) radial growth rate in E. agallocha is significantly lower in the high saline zone, (II) monsoonal precipitation enhances radial growth, (III) large-scale climate drivers, such as sea surface temperature (SST) of the equatorial Pacific and the Indian Ocean, and the El Niño-Southern Oscillation (ENSO) influence tree growth through modifying the local precipitation in varying seasons, and (IV) local habitat variables, such as soil salinity and siltation have combined effects on radial growth of trees although soil salinity exhibits stronger negative effect. Our results highlight the importance of integrating both climate and local habitat variables with tree-ring data for a better understanding of tree growth dynamics under changing environment.

Spatial and Temporal Trends in the Timing of Budburst for Australian Wine Regions

Background and Aims. This research investigates spatial and temporal trends in budburst timing across Australian wine regions from 1910–2019. The potential drivers of these observed trends were then identified, including anthropogenic climate change and large-scale climate drivers (El Nino–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), and Southern Annular Mode (SAM)). Methods and Results. The timing of budburst was approximated using accumulation measurements applied to Australia wide gridded temperature data. We show that the modelled budburst date has been gradually shifting to earlier in the year for most (95%) Australian wine regions, at an average rate of one day every 24 years. This linear trend in budburst timing is likely to be associated with steadily increasing air temperatures due to anthropogenic climate change. Significant interannual variability was also observed and was correlated with IOD and SAM; however, no significant relationship was found with ENSO. Positive IOD phases result in budburst occurring on average four days earlier than the long-term average; however, this can be as high as eight days. Conclusions. The results of this study highlight that budburst timing for wine grapes is not a stationary phenomenon and is influenced by both natural and anthropogenic conditions. Significance of the Study. Understanding variability and trends in modelled budburst timing will assist tactical and strategic management practices and improve phenological modelling and adaptation planning for climate change.

Understanding the variability of heatwave characteristics in southern Africa

Southern Africa is among the regions facing the impacts of weather-related extremes (e.g., droughts, floods and heatwaves). In this study we investigate the spatial and temporal variability of heatwave characteristics (duration, number, frequency, amplitude, and cumulative magnitude) over the southern African region during the extended austral summer (November to March) for the period of 1981/82–2017/2018. Heatwaves were identified using three different approaches, namely, the 90th percentile of minimum daily temperature (CTN90pct), the 90th percentile of daily maximum temperature (CTX90pct) and the Excess Heat Factor (EHF), which is a relatively new method. We performed a Principal Component Analysis (PCA) on each of the five heatwaves characteristics to extract the dominant modes. We inspected the role of large-scale climate drivers on the different modes of heatwave characteristics. The results reveal that all the three heatwave definitions capture the key spatial patterns of heatwave characteristics over the southern African region. However, compared to other two definitions, the minimum temperature-based definition recorded fewer heatwave events over the bulk southern Africa. The El Niño Southern Oscillation (ENSO) was identified as one of the key climate drivers associated with different heatwave characteristics over the southern African region. The findings from this study expand our understanding of heatwaves variability over southern Africa. Moreover, results from this study are relevant to the heatwave forecasting over southern Africa.

Drought over Southeast Asia and Its Association with Large-Scale Drivers

Abstract In this study, the spatiotemporal variability of drought over the entire Southeast Asia (SEA) region and its associations with the large-scale climate drivers during the period 1960–2019 are investigated for the first time. The 12-month Standardized Precipitation Evapotranspiration Index (SPEI) was computed based on the monthly Global Precipitation Climatology Centre (GPCC) precipitation and the monthly Climate Research Unit (CRU) 2-m temperature. The relationships between drought and large-scale climate drivers were examined using the principal component analysis (PCA) and maximum covariance analysis (MCA) techniques. Results showed that the spatiotemporal variability of drought characteristics over SEA is significantly different between mainland Indochina and the Maritime Continent and the difference has been increased substantially in recent decades. Moreover, the entire SEA is divided into four homogeneous drought subregions. Drought over SEA is strongly associated with oceanic and atmospheric large-scale drivers, particularly El Niño–Southern Oscillation (ENSO), following by other remote factors such as the variability of sea surface temperature (SST) over the tropical Atlantic, the Pacific decadal oscillation (PDO), and the Indian Ocean dipole mode (IOD). In addition, there exists an SST anomaly dipole over the Pacific Ocean, which modulates the atmospheric circulations and consequently precipitation over SEA, affecting drought conditions in the study region.

Emerging signals of declining forest resilience under climate change

Forest ecosystems depend on their capacity to withstand and recover from natural and anthropogenic perturbations (that is, their resilience)1. Experimental evidence of sudden increases in tree mortality is raising concerns about variation in forest resilience2, yet little is known about how it is evolving in response to climate change. Here we integrate satellite-based vegetation indices with machine learning to show how forest resilience, quantified in terms of critical slowing down indicators3–5, has changed during the period 2000–2020. We show that tropical, arid and temperate forests are experiencing a significant decline in resilience, probably related to increased water limitations and climate variability. By contrast, boreal forests show divergent local patterns with an average increasing trend in resilience, probably benefiting from warming and CO2 fertilization, which may outweigh the adverse effects of climate change. These patterns emerge consistently in both managed and intact forests, corroborating the existence of common large-scale climate drivers. Reductions in resilience are statistically linked to abrupt declines in forest primary productivity, occurring in response to slow drifting towards a critical resilience threshold. Approximately 23% of intact undisturbed forests, corresponding to 3.32 Pg C of gross primary productivity, have already reached a critical threshold and are experiencing a further degradation in resilience. Together, these signals reveal a widespread decline in the capacity of forests to withstand perturbation that should be accounted for in the design of land-based mitigation and adaptation plans.

Ecological niche models of Welwitschia mirabilis and its subspecies in the Namib desert

Welwitschia mirabilis, an endemic gymnosperm of the Namib Desert, has two subspecies: ssp. namibiana, found in Namibia, and the typical ssp. mirabilis occurring in Angola and Namibia. Despite decades of research little is known about the distribution and ecological preferences of this species. In this paper, we study the ecological niche and the geographical and environmental niche overlaps of the subspecies. We applied ecological niche models using Maxent to an extensive occurrence dataset of W. mirabilis encompassing all known localities of the species. We prepared sets of environmental predictors for climate, vegetation biomass, geology, and topography. Subspecies models were best explained by climate and geology and revealed clearly disjunct ranges. Furthermore, subspecies models were superior to the species model in terms of realism and model quality. This supports the view that infraspecific variability in ecological preferences for Welwitschia mirabilis is high. This conclusion is further supported by the finding that subspecies showed little overlap in geographical and environmental niche space, confirming their unique environmental preferences. Rainfall above 200 mm was a clear threshold for both subspecies although they were separated at about 22°C mean annual temperature. In sum, our study identified the important large-scale climatic drivers and thresholds that define the disjunct ranges of the subspecies and represent the first realistic distribution map of Welwitschia mirabilis.

Large-scale climatic drivers of bison distribution and abundance in North America since the Last Glacial Maximum

As the dominant large herbivore in midcontinent North America since the terminal Pleistocene, bison (Bison spp.) have been a fundamental component of ecosystems and economies. Despite the importance of bison in late Quaternary North America, large-scale (regional to continental) patterns of bison biogeography are not well understood. Here we integrate archaeological and paleontological bison occurrence data with simulated climate data to better understand long-term drivers of bison distribution and abundance in North America. We used these records to model bison distribution and abundance over the past 20 thousand years at 1-thousand-year intervals. Our results show that late Quaternary changes in the distribution and abundance of bison were influenced by large-scale trends in temperature and precipitation. The distribution of bison since the Bølling–Allerød Interstadial (ca. 14 ka) is primarily explained by seasonal temperature patterns (mean temperature of the coldest quarter is the most important variable for 12 of the 14 1-thousand-year intervals). The modeled climate of bison distributions progressively narrowed since the Last Glacial Maximum (ca. 20 ka) as bison populations retracted from disjunct Pleistocene refugia and congregated in midcontinent rangelands. Through the Pleistocene-Holocene transition, bison experienced rapidly warming summer temperatures that increased faster in midcontinent North America than other regions and the continent as a whole. Model results suggest that Holocene bison abundance was influenced by hydroclimatic shifts that affected the quality and availability of forage. Bison abundances decreased through the dry early and mid-Holocene and increased when moisture availability improved in the late Holocene. We infer that bison have thrived under a broad range of environmental conditions since the Last Glacial Maximum and that the climatic and biogeographic space occupied by bison narrowed in recent millennia.

Drying in the low-latitude Atlantic Ocean contributed to terrestrial water storage depletion across Eurasia

Eurasia, home to ~70% of global population, is characterized by (semi-)arid climate. Water scarcity in the mid-latitude Eurasia (MLE) has been exacerbated by a consistent decline in terrestrial water storage (TWS), attributed primarily to human activities. However, the atmospheric mechanisms behind such TWS decline remain unclear. Here, we investigate teleconnections between drying in low-latitude North Atlantic Ocean (LNATO) and TWS depletions across MLE. We elucidate mechanistic linkages and detecte high correlations between decreased TWS in MLE and the decreased precipitation-minus-evapotranspiration (PME) in LNATO. TWS in MLE declines by ~257% during 2003-2017 due to northeastward propagation of PME deficit following two distinct seasonal landfalling routes during January-May and June-January. The same mechanism reduces TWS during 2031-2050 by ~107% and ~447% under scenarios SSP245 and SSP585, respectively. Our findings highlight the risk of increased future water scarcity across MLE caused by large-scale climatic drivers, compounding the impacts of human activities.