Anticipating climate-related crop stress requires crop-risk information that is spatially explicit, probabilistic and fast enough for large seasonal forecast and climate-scenario ensembles. We present the Surrogate Engine for Crop Simulations Framework (SECSF), a deep-learning framework that emulates the process-based ECroPS model using only daily minimum and maximum temperature and precipitation. Trained on ERA5-forced ECroPS simulations for grain maize and spring barley, SECSF closely reproduces daily crop-growth dynamics and harvest timing while reducing computational cost by around four orders of magnitude, enabling ensemble-scale inference suitable for research and operational pipelines such as agricultural early warning and adaptation planning under uncertainty across seasonal-to-climate timescales. When forced with seasonal forecast data, SECSF captures spatially coherent crop-risk patterns across Europe in the high-impact year 2022 and is consistent with independent monitoring, supporting its use for probabilistic Areas of Concern. Under CMIP6 scenarios, SECSF identifies the Mediterranean basin as a hotspot of maize-risk signals through mid-century, with more mixed signals in central and northern Europe.

Crop models have been widely used to optimize nitrogen (N) applications for agronomic decision-making, but uncertainties in model calibration under varying N levels, particularly the effects of phenology choice for calibration, remain underexplored. This study employed the ORYZA v3 model, coupled with a global optimization algorithm, to assess how different calibration strategies affected predictions of leaf area index (LAI) and biomass in two rice varieties under four N levels (0, 90, 180, and 270 kg ha-1). The results indicate when the model is calibrated separately for each N level, predictive accuracy varies considerably for both LAI and biomass, reflecting the difference in crop response to N availability. When calibrating the model simultaneously with multiple N levels from a single phenology phase, variability from different selected phenology phases becomes the dominant source of model uncertainty, rather than N levels. Specifically, calibrations using measured data from the stem elongation to anthesis (SA) and anthesis to maturity (AM) phases across N levels provide the most accurate predictions for LAI (RMSE: 0.51–1.92 m² m-²; R²≥0.88) and biomass (RMSE: 551–2619 kg ha-1; R²≥0.96), respectively. In contrast, calibrations using measured data from early-season (transplanting to stem elongation) result in the least reliable predictions. Combined-phase calibrations using SA and AM phases result in the best predictions for both LAI and biomass, owing to their balanced representation of pre- and post-anthesis growth dynamics. This approach significantly reduces uncertainty from phase selection. However, variability in N application rates emerges as the primary uncertainty source in model simulations, emphasizing the importance of careful selection of N levels in calibration datasets, particularly when measured data span two-thirds of the growing season. These findings offer valuable insights into improved calibration practice for precise N management, highlighting the critical role of both phenology phase and N treatment selection.

We present the Surrogate Engine for Crop Simulations for Maize (SECS4M), a deep-learning emulator designed to replicate the process-based ECroPS crop growth model for grain maize in Europe while enabling computationally efficient, large-scale applications in climate services. SECS4M is built on a nested Long Short-Term Memory architecture capturing short- and long-term weather–crop interactions, while it ingests only three daily meteorological inputs, minimum and maximum temperature and total precipitation, thus minimizing the uncertainty that follows the use of a much wider input stream as in ECroPS. Trained on ERA5-forced yield outputs, SECS4M accurately reproduces crop growth trajectories, harvest timing, and yield distributions. Computational requirements are reduced from ~70s to ~0.008s per grid-cell–year, a four-order-of-magnitude speed-up that enables ensemble-scale, operational use.Forced with bias-adjusted SEAS5.1 forecasts, SECS4M reproduces observed 2022 impacts and supports probabilistic identification of Areas of Concern (AoC) based on tercile-based yield anomalies. Under CMIP6 scenarios SSP3-7.0 and SSP5-8.5 to 2050, the emulator highlights specific regions as persistent hotspots of yield risk, while others exhibit mixed signals. SECS4M thus provides a scalable, digital twins enabled and data-efficient framework for seasonal forecasting, AoC mapping, and scenario analysis. Finally, the methodology can be extended to other crops and can be tested for its potential on other regions.

Many studies have shown that compounding extreme events are likely to exacerbate socio-economic risks compared to single extremes. Despite this important fact, studies focussing on the connectivity of extreme events and their associated impacts frequently have some shortcomings. First, extreme events such as droughts and heat waves are often predefined through thresholds, restricting the class of meteorological events leading to the observed impacts. The choice of threshold for defining these extreme events is also often of meteorological and/or statistical nature and thus potentially unsuitable for the holistic identification of the associated impacts. Furthermore, impacts can arise from combinations of non-extreme events that might fall short of the threshold-based identification, thereby limiting the ability to account for key dynamics that determine the risk associated with compound events. Our study aims to overcome those shortcomings by linking climate events with their observed impacts in agriculture. We analyse wet and warm late winters followed by dry and hot springs, and the associated agricultural damages in Europe with the aim of reconstructing these compound events based on the observed impact. A first analysis is conducted for winter wheat impacts in France, the largest European winter wheat producer. We identify agro-climatic zones based on multivariate time series clustering and employ a regularized generalized canonical correlation analysis to identify the large-scale drivers of crop variability for these regions. The patterns that emerge from the analysis are characterized by wet and warm conditions in January and February linked to a positive North Atlantic Oscillation (NAO) state, followed by warm and dry conditions in April induced by a tripole with a blocking high over Central Europe. Using imbalanced random forests, we construct objective bounds and define thresholds to identify which temperatures are warm enough or which water balances are low enough to be associated with significant effect on crop yield reduction. Our results indicate that imbalanced random forests can predict these types of events reasonably well at the local scale, and that the derived thresholds are mostly lower than the commonly used thresholds for detecting similar extreme events. The latter illustrates that the combination of non-extreme climate events can indeed be detrimental to agricultural production in Europe, which is also crucial as the analysed types of events are predicted to occur more often in the future as a result of climate change.

It is increasingly recognized that the multiple and systemic impacts of Earth system change threaten the prosperity of society through altered land carbon dynamics, freshwater variability, biodiversity loss, and climate extremes. For example, in 2022, there are about 400 climate extremes and natural hazards worldwide, resulting in significant losses of lives and economic damage. Beyond these losses, comprehensive assessment on societal well-being, ecosystem services, and carbon dynamics are often understudied. The rapid expansion of geospatial, atmospheric, and socioeconomic data provides an unprecedented opportunity to develop systemic indices to account for a more comprehensive spectrum of Earth system change risks and to assess their socioeconomic impacts. We propose a novel approach based on the concept of syndromes that can integrate synchronized changes in biosphere, atmosphere, and socioeconomic trajectories into distinct co-evolving phenomena. While the syndrome concept was applied in policy related to environmental conservation, it has not been deciphered from systematic data-driven approaches capable of providing a more comprehensive diagnosis of anthropogenic impacts. By advocating interactive dimensionality reduction approaches, we can identify key interconnected socio-environmental changes as syndromes from big data. We recommend future research tailoring syndromes by incorporating granular data, particularly socio-economic, into dimensionality reduction at different spatio-temporal scales to better diagnose regional-to-global atmospheric and environmental changes that are relevant for socioeconomic changes.

The world has recently witnessed many unprecedented climate disasters, often coinciding with other crises such as pandemics, socio-economic instabilities and ecosystem degradation (closely linked to biodiversity loss). These compound shocks exert profound effects on human, environmental, and economic dimensions, presenting substantial implications also from a financial risk standpoint. Consequently, it becomes imperative to transcend the isolated assessment of individual events and associated risks and progress towards an integrated evaluation of interconnected crises. Compound shocks exhibit characteristics marked by non-linear, intricate, and often unpredictable effects on both society and the economy. Consequently, discerning their impacts cannot be simplified to a mere summation of the effects of their individual shocks. The intricate nonlinearities have the potential to amplify the repercussions of climate-related shocks, presenting considerable challenges to financial stability. Recent advancements in the fields of climate impact modelling, catastrophe risk modeling, machine learning, and macroeconomic modeling hold promise in addressing the existing gaps in modeling compound risks. Our study builds on a survey we conducted among twenty-six central banks and supervisory bodies, revealing a consensus on the crucial importance of considering compound shocks in climate change scenario analyses, specifically pertaining to physical. Leveraging the insights garnered from this survey, we set up a research direction towards integration of compound risks into the development of scenario narratives, storylines and (macro-)economic models capable of effectively capturing compound shocks.

Biodiversity loss and ecosystem degradation could pose a substantial threat to financial stability and the wider economy. Despite scientific evidence of the ongoing ecosystem degradation, methodological and data challenges have so far prevented a detailed assessment of the economic and financial risks. While progress has been made in assessing climate change related risks, our understanding of the linkages between the economy and ecosystem service degradation is still limited. Here we pioneer a nuanced approach to understanding the emerging financial risks of ecosystem change. Using the LPLmL-MAgPIE-SEALS modeling framework, we assess physical, transition and financial risks considering feedbacks from climate change, land use, and degrading ecosystem services. Focusing mainly on the EU, we also assess interconnectedness with other global regions where loss of ecosystem services is more pervasive. Our framework includes climate-sensitive spatially explicitly biophysical data within a partial equilibrium land-system model. Modelled land-use patterns are downscaled to derive fine-scale changes in ecosystem service supply and associated economic feedbacks. We assess various scenarios that build on the existing NGFS (Network for Greening the Financial System) framework. These scenarios range from a degraded world without policy interventions, to an integrated climate-nature scenario, with ambitious policies to mitigate both climate and ecosystem service change. The results indicate diverging biodiversity response based on varying climate and nature policy ambition, emphasizing the need to extend biodiversity safeguarding beyond exclusive reliance on climate mitigation policies. Financial risks are assessed through an analysis of sectoral dependencies on various ecosystem services, laying out the basis for a comprehensive framework that supports informed decision-making facing emerging climate and nature-related risks.

It is common knowledge that climate variability and change have a profound impact on crop production. From the principle that “it is green and it grows” to the assessment of the actual impacts of major weather drivers and their extremes on crop growth through the adoption of agro-management strategies informed by tailored and effective climate services, there is a well documented scientific and operational gap. This work focuses on the development, implementation and testing of an AI-based methodology that aims to reproduce a crop growth model informing on grain maize yield in the European domain. A surrogate AI model based on Bayesian deep learning and inference is compared for its efficiency against the process-based deterministic ECroPS model developed by the Joint Research Centre of the European Commission. The rationale behind this effort is that such mechanistic crop models rely on multiple input meteorological variables and are relatively costly in terms of computing resources and time, crucial aspects for a scalable and widely adopted solution. Such approaches make it possible to run very large ensembles of simulations based, for instance, on ensembles of climate predictions and projections and/or a perturbed parametrization (e.g. on the atmospheric CO2 concentration effects). Our surrogate crop model relies on three weather input variables: daily minimum and maximum temperatures and daily precipitation, where the training was performed with the ECMWF-ERA5 reanalysis.

Crop phenology models are pivotal for simulating crop development, predicting yields and guiding agricultural practices. However, uncertainties exist in simulations due to different model structures and variability in model parameters. Although quantifying these contributions to total variability is often conducted at a site-specific level, few attempts to address this for regional crop modelling using field-calibrated parameters. Our study employs six crop phenology models (APSIM, CERES, GDD, Richardson, Sigmoid and Wang) for simulating maturity timings of three representative rice cultivars using trial data within the Sichuan Basin, China. The Leave-One-Out Cross-Validation (LOOCV) is applied for model calibration with a global parameter optimization algorithm and evaluation. Calibrated models show robust prediction capabilities during LOOCV with R2 2 of 0.68-0.95 and RMSE of 2-4 days, though a larger variance is found for evaluation data than for calibration data. Models calibrated with data from sites having frequent high-temperature (Tmax >= 32 max >= 32 degrees C) episodes tend to have better predictability than without high-temperature episodes. Parameter variability, calibrated with different subsets of each cultivar during LOOCV, is low-to-moderate (mostly CV <= 20 %) except for the Sigmoid models curve steepness parameter. For the early-maturity cultivar, parameter variability is spatially the main uncertainty factor, relating to its greater variability of site-specific calibrated parameter values. For the medium- maturity and late-maturity cultivars, the dominant uncertainty source arises from the interplay between model structures and parameters. Parameter variability notably influences the overall uncertainty more than the model structure variability across the region, except in areas prone to high-temperature extremes where divergent model responses predominate. These findings highlight the cultivar-specific nature of simulation uncertainty, but also the critical need to assess the spatial distribution of uncertainty sources. For parameter uncertainty, a broader conceptualization is essential for more accurate quantifications of uncertainty sources, paving the way for improved ensemble crop modelling, especially at a large spatial scale.

Co-design processes involving the scientific community, practitioners, end users and stakeholders can efficiently characterize harmful weather events during the growing season that potentially result in losses of crop yield and quality. This study builds on the experience of the EU Horizon 2020 project MED-GOLD for grape and olive. The identified agro-climate indicators are extended from the MED-GOLD regions to the entire ones where grape and olive are currently grown in Europe and Turkey, and used to assess climate change impacts with intrinsic adaptation relevance stemming from the co-design process. Before 2000, only a low fraction of the European grape and olive growing areas was exposed to extreme weather events as revealed by the agro-climate indicators, but this has changed rapidly afterward. Projections show increasingly widespread extreme high temperature events from 2020 to 2080. Approximately one-third of grapevine regions and over half of olive cultivation areas are expected to experience extreme drought conditions. Additionally, the frequency of compound extreme events will increase in the future, especially in the Mediterranean region and under the high-end emission scenario RCP8.5. This outcome calls for a new decision-making mindset that embeds expected levels of climate variability and extremes as the “new normal” for grape and olive in Europe. This will facilitate deployment of the required biophysical, economic and policy adaptation tools.

Seasonal phenology forecasts are becoming increasingly demanded by winegrowers and viticulturists. Forecast performance needs to be investigated over space and time before practical applications. We assess seasonal forecast performance (skill, probability and accuracy) in predicting flowering and veraison stages of two representative varieties in Portugal over 1993-2017. The state-of-the-art forecast system ECMWF-SEAS5 pro -vides 7-month seasonal forecasts and is coupled with a locally adapted phenology model. Overall, findings illustrate the dependence of forecast performance on initialization timings, regions and predicting subjects (stages and varieties). Forecast performance improves by delaying the initialization timing and only forecasts initialized on April 1st show better skills than climatology on predicting phenology terciles (early/normal/late). The considerable bias of daily values of seasonal climate predictions can represent the main barrier to accurate forecasts. Better prediction performance is consistently found in Central-Southern regions compared to Northern regions, attributing to an earlier phenology occurrence with a shorter forecast length. Comparable predictive skills between flowering and veraison for both varieties imply better predictability in summer. Consequently, promising seasonal phenology predictions are foreseen in Central-Southern wine regions using forecasts initialized on April 1st with approximately 1-2/3-4 months lead time for flowering/veraison: potential pre -diction errors are similar to 2 weeks, along with an overall moderate forecast skill on categorical events. However, considerable inter-annual variability of forecast performance over the same classified phenology years reflects the substantial influence of climate variability. This may represent the main challenge for reliable forecasts in Mediterranean regions. Recommendations are suggested for methodological innovations and practical applica-tions towards reliable regional phenology forecasts.

IntroductionCassava production is essential for food security in sub-Saharan Africa and serves as a major calorie-intake source in Nigeria. Estimating the yield gap in Nigeria is essential to indicate the most important limiting factors for production, and identify the yield gap hotspot areas. Secondly, these assessments may help set agendas in policy development and research prioritization where current information is scarce. Materials and methodsHere, Wwe used a crop model, LINTUL5, calibrated for five different cassava varieties based on field experiments embedded into a modeling framework SIMPLACE to estimate potential, water- and nutrient (current) limited cassava yield gaps (YG) as affected by climate factors and contributing a better understanding of yield gaps and its potentials in 30 states of Nigeria. ResultsOur study shows that cumulative radiation and precipitation were the most significant factors associated with cassava yield variability (p < 0.01). The YG averaged across states was estimated as 18.2 Ton7ha-1, with a maximum of 31.2 Ton7ha-1 35 in Kano state. Across the states, nutrient limitation accounts for 55.3% of the total cassava yield gap, while the remaining 44.7% is attributed to water limitation. The highest untapped water-limited yields were estimated in the northern states, such as Bauchi, Gombe, and Sokoto, characterized by the short rainy season. ConclusionOur results showed that most northern states are better equipped to become leading cassava producers in Nigeria under adequate crop management practices involving irrigation and soil fertility enhancement. We reached this conclusion because the northern states usually receive the highest radiation from their characteristic reduced cloud cover, even Therefore, policy and management interventions can be prioritized in these areas. Conclusively, the current cassava yield levels can be increased by a factor of five by emphasizing nutrient and soil health management and irrigation, particularly in areas characterized by a shorter rainy season (Sudan Savanna) in Nigeria.

Unfavorable and extreme climate events such as drought and heat stress affect wheat production and food security globally. Predicting such climate events in the next decade is of great interest for decision-makers, as this time horizon coincides with the strategic planning of many stakeholders in the wheat sector. To address this, we assess the forecast quality in predicting the evolution of drought and heat stress conditions using two proxy user-oriented drought and heat stress indicators: Standardized Potential Evapotranspiration Index (SPEI6) and Heat Magnitude Day Index (HMDI3) on a multi-annual timescale (forecast years 1 to 5). In particular, we present the probabilistic skill and reliability of decadal forecast to predict these indices for the months preceding wheat harvest on a global spatial scale. We use decadal forecasts from the Community Earth System Model Decadal Prediction Large Ensemble (CESM-DPLE), which contributes to the Decadal Climate Prediction Project (DCPP) of CMIP6. Following this, we demonstrate the potential applicability of these forecasts to enhance the adaptation and mitigation activities in the wheat sector by presenting the forecast of multi-year averaged SPEI6 and HMDI3 based on categorical events for the period 2016-2020 along with the corresponding observational values.

Extreme climate events can have a significant negative impact on maize productivity, resulting in food scarcity and socioeconomic losses. Thus, quantifying their effect is needed for developing future adaptation and mitigation strategies, especially for countries relying on maize as a staple crop, such as South Africa. While several studies have analyzed the impact of climate extremes on maize yields in South Africa, little is known on the quantitative contribution of combined extreme events to maize yield variability and the causality link of extreme events. This study uses existing stress indices to investigate temporal and spatial patterns of heatwaves, drought, and extreme precipitation during maize growing season between 1986/87 and 2015/16 for South Africa provinces and at national level and quantifies their contribution to yield variability. A causal discovery algorithm was applied to investigate the causal relationship among extreme events. At the province and national levels, heatwaves and extreme precipitation showed no significant trend. However, drought severity increased in several provinces. The modified Combined Stress Index (CSIm) model showed that the maize yield nationwide was associated with drought events (explaining 25% of maize yield variability). Heatwaves has significant influence on maize yield variability (35%) in Free State. In North West province, the maize yield variability (46%) was sensitive to the combination of drought and extreme precipitation. The causal analysis suggests that the occurrence of heatwaves intensified drought, while a causal link between heatwaves and extreme precipitation was not detected. The presented findings provide a deeper insight into the sensitivity of yield data to climate extremes and serve as a basis for future studies on maize yield anomalies.

&lt;p dir=&quot;ltr&quot;&gt;Seasonal climate predictions are becoming important to increase preparedness and adapt agricultural decision making before and within the growing season. Evapotranspiration represents a major component of the water cycle and agricultural water balance, and is thus of key relevance also for assessment of crop growth and irrigation water needs. This study investigates the comparison of different empirical methods for reference evapotranspiration calculation with the reference Penman Monteith method, implemented in the AquaCrop model, over different climate zones of Croatia using national high resolution high quality meteorological dataset in comparison with EOBS and AgERA5 datasets. This research aims to address the appropriateness of using the method with limited data to reduce uncertainty in crop yield modelling with seasonal climate forecasts. &lt;/p&gt; &lt;p dir=&quot;ltr&quot;&gt;The motivation for the research stems from the need to use a reduced number of parameters for ETo calculation, which as such enters the crop yield model. Estimation of seasonal predictions of ETo relies on the skill of prediction of multiple meteorological variables; in the case of Penman Monteith, the estimates would rely on the skill of temperature, vapour pressure, winds speed and global solar radiation prediction. Reducing the number of input variables for estimating the reference evapotranspiration might reduce the overall reliance on the limited skill of seasonal predictions of above-mentioned meteorological variables.&amp;#160;&lt;/p&gt; &lt;p dir=&quot;ltr&quot;&gt;In testing the usefulness of the methods, the interpolated field of the national data network, EOBS dataset, and AgERA5 for the period from 1981 to 2020 are used. Preliminary results show that for the continental area of Croatia the Trajkovic method is satisfactory for ETo calculation, and the EOBS dataset provides better results than AgERA5 when compared to high resolution national dataset. In addition, we analyse the ETo seasonal prediction skill by comparing both methods, original Penman Monteith and Trajkovic method with reduced number of input variables. This will give us an important insight on the relevance of the meteorological dataset choice as well as the ET0 method selection for seasonal prediction purposes. Since there is no comprehensive research on the reference crop evapotranspiration in Croatia, this research fills the gap of knowledge at the local level, but also contributes to the global picture&lt;/p&gt;

&lt;p&gt;The rise of surface temperature due to climate change has a significant impact on the growth of different plant species, including aromatic plants, agricultural crops, fruit trees, and forests. Within the framework of the MICROSERVICES project, that aims on improving the capacity to predict the cascading effects of climate change on microbial diversity, crop-microbiome interactions and agricultural ecosystem functions, we studied the future projection of agro-climatic zones over Europe under two different IPCC Representative Concentration Pathways (RCP4.5 and RCP8.5). To this end, an ensemble of 11 bias-corrected EURO-CORDEX simulations covering the period 1981-2100 was used following the methodology of Ceglar et al. (2019)*. The agro-climatic zones were identified based on two temperature-related parameters, the Growing Season Length (GSL) and the Active Temperature Sum (ATS). The categorization into one of the 8 agro-climatic zones was implemented by applying the k-means clustering method for the reference period 1981-2010. Then, the agro-climatic zone patterns over Europe for the intermediate period 2031-2060 and the end-of-the-century period 2071-2100 were compared against that of the reference period. Our results point towards a strong northward shift of the agroclimatic zones, especially under the no-mitigation emission scenario (RCP8.5) at the end of the century. For the moderate mitigation scenario (RCP4.5), a significant shift of the agroclimatic zones is also observed for the intermediate and the end-of-the-century periods for extended areas in Europe. Our results are in line with that of Ceglar et al. (2019), who studied the migration of agro-climatic zones over Europe under the 2 &lt;sup&gt;o&lt;/sup&gt;C warming level utilizing a subset of the simulations used in this study. *&lt;em&gt;Ceglar et al., Earth&#039;s Future, 7, 1088-1101, https://doi.org/10.1029/2019EF001178, 2019.&lt;/em&gt;&lt;/p&gt;&lt;p&gt;This research was supported by the MICROSERVICES project funded by the General Secretariat for Research and Innovation (GSRI, Greece) under the Action ERANETs 2021A [Call ID: 037KE - A/A MIS 4888] (Project Number: T12ERA5-00075) and through the 2019-2020 BiodivERsA joint call for research proposals [BiodivClim ERA-Net COFUND programme].&lt;/p&gt;

Investigating the global connectivities of extreme events is vital for accurate risk reduction and adaptation planning. While human and natural systems have a certain resilience level against single extremes, they may be unable to cope with multiple extreme events whose impacts tend to be amplified in a non-linear relationship. Concurrent droughts and heatwaves are frequently linked to severe damage in socioeconomic sectors such as agriculture, energy, health, and water resources. They can also have detrimental effects on natural ecosystems. Here, we detect global scale dependencies of large-scale droughts and heatwaves using an AI-enhanced point process-based approach, where large-scale events are defined to occur when a certain amount of grid points (e.g., 20%) of a given region of interest experiences heatwave or drought conditions. The classic inhomogeneous and non-stationary J-function can determine whether the occurrence of the events shows clustering, inhibition or independence. However, the analysis and interpretation of this function are usually affected by a high degree of subjectiveness, and its application for large datasets and/or ensembles is challenging. The proposed AI-based automated interpretation tool replaces a subjective and user-dependent approach. Monte Carlo simulations based on standard point process models, reflecting the aforementioned dependence structures, are utilized, allowing the dependence structure to be labeled and the classification problem to be trained using Deep Learning algorithms. To identify the global connectivities of large-scale droughts and heatwaves, we first detect extreme events at the grid scale based on appropriately selected indices. A cluster analysis pinpoints areas with similar drought and heatwave patterns, thus identifying the regions of interest for the large-scale events. For these events we compute the J-functions, and the dependence structure of the large-scale events is then classified by the AI-tool. Links to teleconnections (such as the El Niño-Southern Oscillation and the North Atlantic Oscillation) can be further identified by analyzing the dependencies conditioning on the teleconnection phase under consideration. The proposed tool can be used in diverse research questions where a point process approach is appropriate, and thus has applications beyond climate science.

We investigate future changes of agro-climatic zones over Europe under two different IPCC Representative Concentration Pathways (RCP4.5 and RCP8.5) based on an ensemble of 11 bias-corrected regional climate model simulations covering the period 1981–2100. Eight distinct agro-climatic zones were identified by applying the k-means clustering method for the reference period 1981–2010 on two temperature-related parameters: the growing season length and the active temperature sum. Our results point towards a strong northward shift of the agro-climatic zones, especially under RCP8.5, towards the end of the century. A significant shift of the agro-climatic zones is also projected for RCP4.5 in the near-future and the end-of-the-century periods for large areas in Europe.

&lt;p&gt;A growing literature investigates the dynamic impacts of climate shocks on economy. Drought and heat stress can lead to substantial degradation of human health, agriculture and food production, energy sector, transport and mobility, and other sectors. These impacts are manifested in the economy and financial systems through different transmission channels and can result in increased costs and economic losses, with cascading effects to financial stability as well. Drought and heat waves often compound, with their interconnected risks being complex, challenging to understand and manage. Moreover, the impacts of compound shocks can be significantly larger than the sum of the individual shocks, mainly due to non-linearity of impacts. To address this issue, we investigate in this study to what extent the effect of extreme temperature variations on key economic outputs depends on the occurrence of droughts and high precipitation in the recent past. We estimate a novel Bayesian Smooth Transition Vector Autoregression, allowing the data to inform the pattern in which the model moves between extreme droughts and extreme wet conditions. We find that the occurrence of droughts within the last 12 months has important consequences for the way in which temperature shocks affect aggregate prices. These effects vary to some degree across countries as well as across sectors.&lt;/p&gt;