Search

Global hydrological models (GHMs) are widely used to assess the impact of climate change on streamflow, floods, and hydrological droughts. For the ‘model evaluation and impact attribution’ part of the current round of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP3a), modelling teams generated historical simulations based on observed climate and direct human forcings with updated model versions. Here we provide a comprehensive evaluation of daily and maximum annual discharge based on ISIMIP3a simulations from nine GHMs by comparing the simulations to observational data from 644 river gauge stations. We also assess low flows and the effects of different river routing schemes. We find that models can reproduce variability in daily and maximum annual discharge, but tend to overestimate both quantities, as well as low flows. Models perform better at stations in wetter areas and at lower elevations. Discharge routed with the river routing model CaMa-Flood can improve the performance of some models, but for others, variability is overestimated, leading to reduced model performance. This study indicates that areas for future model development include improving the simulation of processes in arid regions and cold dynamics at high elevations. We further suggest that studies attributing observed changes in discharge to historical climate change using the current model ensemble will be most meaningful in humid areas, at low elevations, and in places with a regular seasonal discharge as these are the regions where the underlying dynamics seem to be best represented.

Disturbances in vegetated land could dramatically affect the process of vegetation growth and reshape the land cover state. The overall greenup of vegetation on the Tibetan Plateau (TP) has almost served as a consensus to date. However, we still lack consistent acquisitions on the timing, the spatial patterns, and the temporal frequency of vegetation disturbance over the TP, limiting the capacity for planning land management strategies. Therefore, we explored the spatiotemporal pattern and variation of vegetation disturbances across the TP during the past decades and analyzed the disturbance agents. We utilized 37-year Landsat time series images and field observations coupled with a temporal segmentation approach to characterize the spatiotemporal pattern of vegetation disturbances across the TP for the period 1986–2018. The results from this study revealed that 75.71 M ha (accounting for 29.34% of TP’s area) vegetation area underwent at least one disturbance, of which 8.44 M ha area ever experienced large-scale disturbances (disturbance area greater than 0.9 ha and disturbance magnitude (the difference between the spectral value of pre-disturbance and that of post-disturbance) over 0.2). Further, the spatial distributions of these large-scale disturbances varied over time: before 2002, the disturbed sites were evenly distributed over the southeast part of the TP probably induced by overgrazing and unscientific livestock management, while after 2002, most disturbances were concentrated in the south of the Yarlung Tsangpo, mainly caused by anthropogenic activities, such as urban area, roadways, railway, and water control projects. This study presents an effort to characterize vegetation disturbances and their variations over the past decades on the TP, which provides crucial insights toward a complete understanding of vegetation dynamics and its causal relationship with human activities.

Amplified climate warming has led to permafrost degradation and a shortening of the winter season, both impacting cost-effective overland travel across the Arctic. Here we use, for the first time, four state-of-the-art Land Surface Models that explicitly consider ground freezing states, forced by a subset of bias-adjusted CMIP5 General Circulation Models to estimate the impact of different global warming scenarios (RCP2.6, 6.0, 8.5) on two modes of winter travel: overland travel days (OTDs) and ice road construction days (IRCDs). We show that OTDs decrease by on average −13% in the near future (2021–2050) and between −15% (RCP2.6) and −40% (RCP8.5) in the far future (2070–2099) compared to the reference period (1971–2000) when 173 d yr ^−1 are simulated across the Pan-Arctic. Regionally, we identified Eastern Siberia (Sakha (Yakutia), Khabarovsk Krai, Magadan Oblast) to be most resilient to climate change, while Alaska (USA), the Northwestern Russian regions (Yamalo, Arkhangelsk Oblast, Nenets, Komi, Khanty-Mansiy), Northern Europe and Chukotka are highly vulnerable. The change in OTDs is most pronounced during the shoulder season, particularly in autumn. The IRCDs reduce on average twice as much as the OTDs under all climate scenarios resulting in shorter operational duration. The results of the low-end global warming scenario (RCP2.6) emphasize that stringent climate mitigation policies have the potential to reduce the impact of climate change on winter mobility in the second half of the 21st century. Nevertheless, even under RCP2.6, our results suggest substantially reduced winter overland travel implying a severe threat to livelihoods of remote communities and increasing costs for resource exploration and transport across the Arctic.

Climate change is projected to increase the aridity of semi-arid ecosystems, including Mongolian grasslands (MG), which provide ecosystem services that support food supply and pastoralist lifestyle. Here, we conducted a grid-scale (0.5° × 0.5°) probabilistic risk assessment of MG under climate change for 40 years (1976–2015) based on probability theory. We evaluated changes of risk (impacts) and vulnerability of MG to drought between the recent two decades R20 = 1996–2015 and the previous two decades P20 = 1976–1995. The risk is quantified as the product of the probability of hazardous drought and ecosystem vulnerability. The probability of hazardous drought is defined from the Standardized Precipitation–Evapotranspiration Index. Vulnerability is defined as the expected differences of key ecosystem variables between years with and without hazardous conditions. The ecosystem variables are productivity (peak aboveground biomass, net primary productivity, and leaf area index) and root-zone plant-available soil moisture, simulated with a process-based vegetation model Organizing Carbon and Hydrology in Dynamic Ecosystems-Grassland Management validated with field observations of biomass and soil moisture. Results reveal that MG experienced more frequent hazardous droughts with rapid warming and slight drying during R20 aggravated by ever-increasing grazing intensity (34% compared to P20), which resulted in a reduction in soil water availability and grassland productivity, particularly in northeastern areas (20%–65%). The risk of drought to productivity increased by 10% between P20 and R20 over extended areas, particularly in northcentral and northeast Mongolia. The increase in the risk to MG was mainly caused by climate change-induced increase in the probability of hazardous drought and, to a lesser extent, by the increasing vulnerability. Recent droughts modify the risk to grasslands, particularly in northcentral and northeast Mongolia, suggesting that these regions need strategic management for both adaptation and ecosystem conservation to cope with climate change impacts.

Global terrestrial vegetation is greening, particularly in mountain areas, providing strong feedbacks to a series of ecosystem processes. This greening has been primarily attributed to climate change. However, the spatial variability and magnitude of such greening do not synchronize with those of climate change in mountain areas. By integrating two data sets of satellite-derived normalized difference vegetation index (NDVI) values, which are indicators of vegetation greenness, in the period 1982–2015 across the Tibetan Plateau (TP), we test the hypothesis that climate-change-induced greening is regulated by terrain, baseline climate and soil properties. We find a widespread greening trend over 91% of the TP vegetated areas, with an average greening rate (i.e. increase in NDVI) of 0.011 per decade. The linear mixed-effects model suggests that climate change alone can explain only 26% of the variation in the observed greening. Additionally, 58% of the variability can be explained by the combination of the mountainous characteristics of terrain, baseline climate and soil properties, and 32% of this variability was explained by terrain. Path analysis identified the interconnections of climate change, terrain, baseline climate and soil in determining greening. Our results demonstrate the important role of mountainous effects in greening in response to climate change.

Arctic ecosystems are particularly vulnerable to climate change because of Arctic amplification. Here, we assessed the climatic impacts of low-end, 1.5 °C, and 2.0 °C global temperature increases above pre-industrial levels, on the warming of terrestrial ecosystems in northern high latitudes (NHL, above 60 °N including pan-Arctic tundra and boreal forests) under the framework of the Inter-Sectoral Impact Model Intercomparison Project phase 2b protocol. We analyzed the simulated changes of net primary productivity, vegetation biomass, and soil carbon stocks of eight ecosystem models that were forced by the projections of four global climate models and two atmospheric greenhouse gas pathways (RCP2.6 and RCP6.0). Our results showed that considerable impacts on ecosystem carbon budgets, particularly primary productivity and vegetation biomass, are very likely to occur in the NHL areas. The models agreed on increases in primary productivity and biomass accumulation, despite considerable inter-model and inter-scenario differences in the magnitudes of the responses. The inter-model variability highlighted the inadequacies of the present models, which fail to consider important components such as permafrost and wildfire. The simulated impacts were attributable primarily to the rapid temperature increases in the NHL and the greater sensitivity of northern vegetation to warming, which contrasted with the less pronounced responses of soil carbon stocks. The simulated increases of vegetation biomass by 30–60 Pg C in this century have implications for climate policy such as the Paris Agreement. Comparison between the results at two warming levels showed the effectiveness of emission reductions in ameliorating the impacts and revealed unavoidable impacts for which adaptation options are urgently needed in the NHL ecosystems.

Actual land evapotranspiration (ET) is a key component of the global hydrological cycle and an essential variable determining the evolution of hydrological extreme events under different climate change scenarios. However, recently available ET products show persistent uncertainties that are impeding a precise attribution of human-induced climate change. Here, we aim at comparing a range of independent global monthly land ET estimates with historical model simulations from the global water, agriculture, and biomes sectors participating in the second phase of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP2a). Among the independent estimates, we use the EartH2Observe Tier-1 dataset (E2O), two commonly used reanalyses, a pre-compiled ensemble product (LandFlux-EVAL), and an updated collection of recently published datasets that algorithmically derive ET from observations or observations-based estimates (diagnostic datasets). A cluster analysis is applied in order to identify spatio-temporal differences among all datasets and to thus identify factors that dominate overall uncertainties. The clustering is controlled by several factors including the model choice, the meteorological forcing used to drive the assessed models, the data category (models participating in the different sectors of ISIMIP2a, E2O models, diagnostic estimates, reanalysis-based estimates or composite products), the ET scheme, and the number of soil layers in the models. By using these factors to explain spatial and spatio-temporal variabilities in ET, we find that the model choice mostly dominates (24%–40% of variance explained), except for spatio-temporal patterns of total ET, where the forcing explains the largest fraction of the variance (29%). The most dominant clusters of datasets are further compared with individual diagnostic and reanalysis-based estimates to assess their representation of selected heat waves and droughts in the Great Plains, Central Europe and western Russia. Although most of the ET estimates capture these extreme events, the generally large spread among the entire ensemble indicates substantial uncertainties.

This paper evaluates the ability of eight global vegetation models to reproduce recent trends and inter-annual variability of biomass in natural terrestrial ecosystems. For the purpose of this evaluation, the simulated trajectories of biomass are expressed in terms of the relative rate of change in biomass (RRB), defined as the deviation of the actual rate of biomass turnover from its equilibrium counterpart. Cumulative changes in RRB explain long-term changes in biomass pools. RRB simulated by the global vegetation models is compared with its observational equivalent, derived from vegetation optical depth reconstructions of above-ground biomass (AGB) over the period 1993–2010. According to the RRB analysis, the rate of global biomass growth described by the ensemble of simulations substantially exceeds the observation. The observed fluctuations of global RRB are significantly correlated with El Niño Southern Oscillation events (ENSO), but only some of the simulations reproduce this correlation. However, the ENSO sensitivity of RRB in the tropics is not significant in the observation, while it is in some of the simulations. This mismatch points to an important limitation of the observed AGB reconstruction to capture biomass variations in tropical forests. Important discrepancies in RRB were also identified at the regional scale, in the tropical forests of Amazonia and Central Africa, as well as in the boreal forests of north-western America, western and central Siberia. In each of these regions, the RRBs derived from the simulations were analyzed in connection with underlying differences in net primary productivity and biomass turnover rate ̶as a basis for exploring in how far differences in simulated changes in biomass are attributed to the response of the carbon uptake to CO _2 increments, as well as to the model representation of factors affecting the rates of mortality and turnover of foliage and roots. Overall, our findings stress the usefulness of using RRB to evaluate complex vegetation models and highlight the importance of conducting further evaluations of both the actual rate of biomass turnover and its equilibrium counterpart, with special focus on their background values and sources of variation. In turn, this task would require the availability of more accurate multi-year observational data of biomass and net primary productivity for natural ecosystems, as well as detailed and updated information on land-cover classification.

Many studies on drought consider precipitation and potential evapotranspiration (PET) impacts. However, catchment water retention is a factor affecting the interception of precipitation and slowing down runoff which also plays a critical role in determining the risks of hydrological drought. The Budyko framework links retention to the partitioning of precipitation into runoff or evapotranspiration. Applied worldwide, we demonstrate that retention changes are the dominant contribution to measured runoff changes in 21 of 33 major catchments. Similarly, assessing climate simulations for the historical period suggests that models substantially underestimate observed runoff changes due to unrepresented water management processes. Climate models show that water retention (without direct water management) generally decreases by the end of the 21st century, except in dry central Asia and northwestern China. Such decreases raise runoff, mainly driven by precipitation intensity increases (RCP4.5 scenario) and additionally by CO _2 -induced stomata closure (RCP8.5). This mitigates runoff deficits (generally from raised PET under warming) by increasing global mean runoff from −2.77 mm yr ^−1 to +3.81 mm yr ^−1 (RCP4.5), and −6.98 mm yr ^−1 to +5.11 mm yr ^−1 (RCP8.5).

The purpose of this study is to evaluate the eight ISIMIP2a biome models against independent estimates of long-term net carbon fluxes (i.e. Net Biome Productivity, NBP) over terrestrial ecosystems for the recent four decades (1971–2010). We evaluate modeled global NBP against 1) the updated global residual land sink (RLS) plus land use emissions ( E _LUC ) from the Global Carbon Project (GCP), presented as R + L in this study by Le Quéré et al ( 2015 ), and 2) the land CO _2 fluxes from two atmospheric inversion systems: Jena CarboScope s81_v3.8 and CAMS v15r2, referred to as F _Jena and F _CAMS respectively. The model ensemble-mean NBP (that includes seven models with land-use change) is higher than but within the uncertainty of R + L, while the simulated positive NBP trend over the last 30 yr is lower than that from R + L and from the two inversion systems. ISIMIP2a biome models well capture the interannual variation of global net terrestrial ecosystem carbon fluxes. Tropical NBP represents 31 ± 17% of global total NBP during the past decades, and the year-to-year variation of tropical NBP contributes most of the interannual variation of global NBP. According to the models, increasing Net Primary Productivity (NPP) was the main cause for the generally increasing NBP. Significant global NBP anomalies from the long-term mean between the two phases of El Niño Southern Oscillation (ENSO) events are simulated by all models ( p < 0.05), which is consistent with the R + L estimate ( p = 0.06), also mainly attributed to NPP anomalies, rather than to changes in heterotrophic respiration (Rh). The global NPP and NBP anomalies during ENSO events are dominated by their anomalies in tropical regions impacted by tropical climate variability. Multiple regressions between R + L, F _Jena and F _CAMS interannual variations and tropical climate variations reveal a significant negative response of global net terrestrial ecosystem carbon fluxes to tropical mean annual temperature variation, and a non-significant response to tropical annual precipitation variation. According to the models, tropical precipitation is a more important driver, suggesting that some models do not capture the roles of precipitation and temperature changes adequately.

Global hydrological models (GHMs) have been applied to assess global flood hazards, but their capacity to capture the timing and amplitude of peak river discharge—which is crucial in flood simulations—has traditionally not been the focus of examination. Here we evaluate to what degree the choice of river routing scheme affects simulations of peak discharge and may help to provide better agreement with observations. To this end we use runoff and discharge simulations of nine GHMs forced by observational climate data (1971–2010) within the ISIMIP2a project. The runoff simulations were used as input for the global river routing model CaMa-Flood. The simulated daily discharge was compared to the discharge generated by each GHM using its native river routing scheme. For each GHM both versions of simulated discharge were compared to monthly and daily discharge observations from 1701 GRDC stations as a benchmark. CaMa-Flood routing shows a general reduction of peak river discharge and a delay of about two to three weeks in its occurrence, likely induced by the buffering capacity of floodplain reservoirs. For a majority of river basins, discharge produced by CaMa-Flood resulted in a better agreement with observations. In particular, maximum daily discharge was adjusted, with a multi-model averaged reduction in bias over about 2/3 of the analysed basin area. The increase in agreement was obtained in both managed and near-natural basins. Overall, this study demonstrates the importance of routing scheme choice in peak discharge simulation, where CaMa-Flood routing accounts for floodplain storage and backwater effects that are not represented in most GHMs. Our study provides important hints that an explicit parameterisation of these processes may be essential in future impact studies.

Terrestrial gross primary productivity (GPP) is the largest component of the global carbon cycle and a key process for understanding land ecosystems dynamics. In this study, we used GPP estimates from a combination of eight global biome models participating in the Inter-Sectoral Impact-Model Intercomparison Project phase 2a (ISIMIP2a), the Moderate Resolution Spectroradiometer (MODIS) GPP product, and a data-driven product (Model Tree Ensemble, MTE) to study the spatiotemporal variability of GPP at the regional and global levels. We found the 2000–2010 total global GPP estimated from the model ensemble to be 117 ± 13 Pg C yr ^−1 (mean ± 1 standard deviation), which was higher than MODIS (112 Pg C yr ^−1 ), and close to the MTE (120 Pg C yr ^−1 ). The spatial patterns of MODIS, MTE and ISIMIP2a GPP generally agree well, but their temporal trends are different, and the seasonality and inter-annual variability of GPP at the regional and global levels are not completely consistent. For the model ensemble, Tropical Latin America contributes the most to global GPP, Asian regions contribute the most to the global GPP trend, the Northern Hemisphere regions dominate the global GPP seasonal variations, and Oceania is likely the largest contributor to inter-annual variability of global GPP. However, we observed large uncertainties across the eight ISIMIP2a models, which are probably due to the differences in the formulation of underlying photosynthetic processes. The results of this study are useful in understanding the contributions of different regions to global GPP and its spatiotemporal variability, how the model- and observational-based GPP estimates differ from each other in time and space, and the relative strength of the eight models. Our results also highlight the models’ ability to capture the seasonality of GPP that are essential for understanding the inter-annual and seasonal variability of GPP as a major component of the carbon cycle.

Simulating vegetation photosynthetic productivity (or gross primary production, GPP) is a critical feature of the biome models used for impact assessments of climate change. We conducted a benchmarking of global GPP simulated by eight biome models participating in the second phase of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP2a) with four meteorological forcing datasets (30 simulations), using independent GPP estimates and recent satellite data of solar-induced chlorophyll fluorescence as a proxy of GPP. The simulated global terrestrial GPP ranged from 98 to 141 Pg C yr ^−1 (1981–2000 mean); considerable inter-model and inter-data differences were found. Major features of spatial distribution and seasonal change of GPP were captured by each model, showing good agreement with the benchmarking data. All simulations showed incremental trends of annual GPP, seasonal-cycle amplitude, radiation-use efficiency, and water-use efficiency, mainly caused by the CO _2 fertilization effect. The incremental slopes were higher than those obtained by remote sensing studies, but comparable with those by recent atmospheric observation. Apparent differences were found in the relationship between GPP and incoming solar radiation, for which forcing data differed considerably. The simulated GPP trends co-varied with a vegetation structural parameter, leaf area index, at model-dependent strengths, implying the importance of constraining canopy properties. In terms of extreme events, GPP anomalies associated with a historical El Niño event and large volcanic eruption were not consistently simulated in the model experiments due to deficiencies in both forcing data and parameterized environmental responsiveness. Although the benchmarking demonstrated the overall advancement of contemporary biome models, further refinements are required, for example, for solar radiation data and vegetation canopy schemes.