Search

Abstract Scientific inquiry has long relied on deterministic algorithms for systematic problem-solving and predictability. However, the rise of artificial intelligence (AI) has revolutionized data analysis, allowing us to uncover complex patterns in large datasets. In this study, we combine these two approaches by using AI to improve the reconstruction of past precipitation events, which is crucial for understanding climate change. Our objective is to leverage AI to map large-scale atmospheric proxies from the ERA5 climate reanalysis and multi-satellite historical precipitation data from the NASA-IMERG GPM constellation to observed precipitation, enhancing the accuracy and the resolution of climate reanalysis. Accurate climate reanalyses are essential, as they provide the most realistic representations of past atmospheric conditions, serving as benchmarks against which climate models are validated. Our AI-enhanced method offers a more accurate and computationally efficient solution compared to deterministic high-resolution precipitation downscaling methods. Additionally, it shows the capability to generalize predictions to new, previously unobserved locations, making it applicable across various regions. By integrating AI with traditional reanalysis techniques, we open up new opportunities for climate science and geosciences, with the potential to improve the accuracy and reliability of climate data, contributing to a better understanding of climate dynamics.

Scientific inquiry has long relied on deterministic algorithms for systematic problem-solving and predictability. However, the rise of artificial intelligence (AI) has revolutionized data analysis, allowing us to uncover complex patterns in large datasets. In this study, we combine these two approaches by using AI to improve the reconstruction of past precipitation events, which is crucial for understanding climate change. Our objective is to leverage AI to map large-scale atmospheric proxies from the ERA5 climate reanalysis and multi-satellite historical precipitation data from the NASA-IMERG GPM constellation to observed precipitation, enhancing the accuracy and the resolution of climate reanalysis. Accurate climate reanalyses are essential, as they provide the most realistic representations of past atmospheric conditions, serving as benchmarks against which climate models are validated. Our AI-enhanced method offers a more accurate and computationally efficient solution compared to deterministic high-resolution precipitation downscaling methods. Additionally, it shows the capability to generalize predictions to new, previously unobserved locations, making it applicable across various regions. By integrating AI with traditional reanalysis techniques, we open up new opportunities for climate science and geosciences, with the potential to improve the accuracy and reliability of climate data, contributing to a better understanding of climate dynamics. [ABSTRACT FROM AUTHOR]

Changes in land use and land cover (LULC) have significant implications for biodiversity, ecosystem functioning, and deforestation. Modeling LULC changes is crucial to understanding anthropogenic impacts on environmental conservation and ecosystem services. Although previous studies have focused on predicting future changes, there is a growing need to determine past scenarios using new assessment tools. This study proposes a methodology for LULC past scenario generation based on transition analysis. Aiming to hindcast LULC scenario in 1970 based on the transition analysis of the past 35 years (from 1985 to 2020), two machine learning algorithms, multilayer perceptron (MLP) and similarity weighted (SimWeight), were employed to determine the driver variables most related to conversions in LULC and to simulate the past. The study focused on the Aristida spp. grasslands in the Uruguayan savannas, where native grasslands have been extensively converted to agricultural areas. Land use and land cover data from the MapBiomas project were integrated with spatial variables such as altimetry, slope, pedology, and linear distances from rivers, roads, urban areas, agriculture, forest, forestry, and native grasslands. The accuracy of the predicted maps was assessed through stratified random sampling of reference images from the Multispectral Scanner (MSS) sensor. The results demonstrate a reduction of approximately 659 934 ha of native grasslands in the study area between 1985 and 2020, directly proportional to the increase in cultivable areas. The MLP algorithm exhibited moderate performance, with notable errors in classifying agriculture and grassland areas. In contrast, the SimWeight algorithm displayed better accuracy, particularly in distinguishing grassland and agriculture classes. The modeled map using SimWeight accurately represented the transitions between grassland and agriculture with a high level of agreement. By modeling the 1970s scenario using the SimWeight model, it was estimated that the Aristida spp. grasslands experienced a substantial reduction in grassland coverage, ranging from 9982.31 to 10 022.32 km2 between 1970 and 2020. This represents a range of 60.8%–61.07% of the total grassland area in 1970. These findings provide valuable insights into the driving factors behind land use change in the Aristida spp. grasslands and offer useful information for land management, conservation, and sustainable development in the region. The study's main contribution lies in the hindcasting of past LULC scenarios, utilizing a tool used primarily for forecasting future scenarios. Integr Environ Assess Manag 2024;20:1140–1155. © 2023 SETAC Key Points: The study's main contribution lies in the hindcasting of past land use and land cover scenarios, using a tool primarily used for forecasting future scenarios.The algorithm's performance is significantly influenced by the heterogeneity of the area of interest studied.We estimate that Aristida spp. fields have lost more than 60% of their original cover in the past 50 years. [ABSTRACT FROM AUTHOR]

Climate warming is expected to increase global methane (CH4) emissions from wetland ecosystems. Although in situ eddy covariance (EC) measurements at ecosystem scales can potentially detect CH4 flux changes, most EC systems have only a few years of data collected, so temporal trends in CH4 remain uncertain. Here, we use established drivers to hindcast changes in CH4 fluxes (FCH4) since the early 1980s. We trained a machine learning (ML) model on CH4 flux measurements from 22 [methane‐producing sites] in wetland, upland, and lake sites of the FLUXNET‐CH4 database with at least two full years of measurements across temperate and boreal biomes. The gradient boosting decision tree ML model then hindcasted daily FCH4 over 1981–2018 using meteorological reanalysis data. We found that, mainly driven by rising temperature, half of the sites (n = 11) showed significant increases in annual, seasonal, and extreme FCH4, with increases in FCH4 of ca. 10% or higher found in the fall from 1981–1989 to 2010–2018. The annual trends were driven by increases during summer and fall, particularly at high‐CH4‐emitting fen sites dominated by aerenchymatous plants. We also found that the distribution of days of extremely high FCH4 (defined according to the 95th percentile of the daily FCH4 values over a reference period) have become more frequent during the last four decades and currently account for 10–40% of the total seasonal fluxes. The share of extreme FCH4 days in the total seasonal fluxes was greatest in winter for boreal/taiga sites and in spring for temperate sites, which highlights the increasing importance of the non‐growing seasons in annual budgets. Our results shed light on the effects of climate warming on wetlands, which appears to be extending the CH4 emission seasons and boosting extreme emissions. [ABSTRACT FROM AUTHOR]

Freshwater mussels are globally imperilled, which is attributed to their sensitivity to changes in streamflow and temperature. The Black River in south‐eastern New Mexico, USA, harbours a stronghold population of the federally endangered unionid Popenaias popeii (Texas hornshell). Decreasing discharge and elevated water temperatures are considered the primary factors responsible for the species' decline. However, the impact of these factors has not been quantified directly, which may hinder conservation efforts.The upper thermal tolerances (LT05 and LT50) of Black River P. popeii larvae (glochidia) and newly transformed juveniles were evaluated. Individuals were acclimated to 27°C and then immersed at five experimental temperatures (28, 30, 32, 34 and 36°C) for 24 h (glochidia) or 96 h (juveniles). The thermal tolerances of P. popeii and its presumed host fishes were then overlain onto in situ water temperature and discharge data to determine thermal exceedances in the past 15 years. Temperature was hindcast back to 2007 using machine learning (random forest model) and coupled with discharge.For glochidia, LT05 was exceeded frequently (41%) and LT50 was exceeded occasionally (13%) during spring and summer periods, whereas juvenile thresholds were never exceeded. Upper continuous duration above threshold (UCAT) analysis revealed periods of catastrophic high temperature, which were used to determine discharge bottlenecks.This study demonstrates how laboratory‐derived physiological thresholds can be used in conjunction with environmental data to evaluate the hydrological needs of aquatic organisms, which is useful in efforts to maintain flow regimes that protect native ecosystems. [ABSTRACT FROM AUTHOR]

Prediction is frequently asserted to be the sine qua non of science, but prediction means different things to different people in different contexts. I organize and explain this diversity by addressing five questions. What does it mean to predict something? To answer this question, I describe concepts of prediction as prophecy, diagnosis, history, repeatability, and fate. What are we trying to predict? Here, I describe how predictions vary along several axes: general to specific, qualitative to quantitative, relative to absolute, point to range, and continuous to discontinuous. Where do predictions come from? In this case, I focus on deductive versus inductive reasoning. How do we test predictions? The answer here is not straightforward and I discuss various approaches and difficulties. How good are predictions? Not surprisingly, it depends on what is being predicted and how we judge success. Importantly, I do not espouse a "best" way to approach prediction but, rather, I outline its diverse manifestations so as to help organize practical thinking on the topic. [ABSTRACT FROM AUTHOR]

The accurate estimation of wind speed is crucial for wind energy production, given the exponential relationship between wind and power. However, this is a challenging task due to the stochastic nature of meteorology. In this study, the Weather Research and Forecasting Model (WRF) with a 9 km spatial resolution was used to simulate hourly wind speed values for Türkiye, using Global Data Assimilation System (GDAS) 0.25° boundary data. Several statistical metrics, including Root Mean Squared Error (RMSE), Mean Bias Error (MB), Index of Agreement (IoA), and Pearson correlation, were used to evaluate the performance of the WRF model. The WRF model, which used CONUS parametrization and was supplied with GDAS boundary data every 6 h, operated for 17,520 h in a 1-month consecutive run. The ANN model, which has Hecht-Nielsen (2n + 1) topology, was used to perform hindcasting of the WRF model. The input layer of the ANN model used temperature, pressure, and wind speed values obtained from WRF. The analysis was done spatio-temporally for 2 years and presented with seasonal and annual performance values. After applying the ANN model to the WRF model, which had initial values of MB of 1.42, RMSE of 2.26, R of 0.51, and IOA value of 0.02, the new MB, RMSE, R, and IOA values were found to be 0.04, 0.96, 0.56, and 0.60, respectively. Therefore, it can be concluded that the ANN model improved the WRF model's wind speed prediction performance in Türkiye by 11% on average, relatively. [ABSTRACT FROM AUTHOR]

Long records of wave parameters are central to the estimation of coastal flooding risk and the causes of coastal erosion. This paper leverages the predictive power of wave height history and correlations with wind speed and direction to build statistical models for time series of wave heights to develop a method to fill data-gaps and extend the record length of coastal wave observations. A threshold regression model is built where the threshold parameter, based on lagged wind speed, explains the nonlinear associations, and the lagged predictors in the model are based on a well-established empirical wind-wave relationship. The predictive model is completed by addressing the residual conditional heteroscedasticity using a GARCH model. This comprehensive model is trained on time series data from 2005 to 2013, using wave height and wind data both observed from a buoy in Long Island Sound. Subsequently, replacing wind data with observations from a nearby coastal station provides a similar level of predictive accuracy. This approach can be used to hindcast wave heights for past decades given only wind information at a coastal station. These hindcasts are used as a representative of the unobserved past to carry out extreme value analysis by fitting Generalized Pareto (GP) distribution in a peaks over threshold (POT) framework. By analyzing longer periods of data, we can obtain reliable return value estimates to help design better coastal protection structures. [ABSTRACT FROM AUTHOR]

The reconstruction or prediction of meteorological records through the Analog Ensemble (AnEn) method is very efficient when the number of predictor time series is small. Thus, in order to take advantage of the richness and diversity of information contained in a large number of predictors, it is necessary to reduce their dimensions. This study presents methods to accomplish such reduction, allowing the use of a high number of predictor variables. In particular, the techniques of Principal Component Analysis (PCA) and Partial Least Squares (PLS) are used to reduce the dimension of the predictor dataset without loss of essential information. The combination of the AnEn and PLS techniques results in a very efficient hybrid method (PLSAnEn) for reconstructing or forecasting unstable meteorological variables, such as wind speed. This hybrid method is computationally demanding but its performance can be improved via parallelization or the introduction of variants in which all possible analogs are previously clustered. The multivariate linear regression methods used on the new variables resulting from the PCA or PLS techniques also proved to be efficient, especially for the prediction of meteorological variables without local oscillations, such as the pressure. [ABSTRACT FROM AUTHOR]

The distribution and transmission of Yersinia pestis, the bacterial agent of plague, responds dynamically to climate, both within wildlife reservoirs and human populations. The exact mechanisms mediating plague's response to climate are still poorly understood, particularly across large environmentally heterogeneous regions encompassing several reservoir species. A heterogeneous response to precipitation was observed in plague intensity across northern and southern China during the Third Pandemic. This has been attributed to the response of reservoir species in each region. We use environmental niche modelling and hindcasting methods to test the response of a broad range of reservoir species to precipitation. We find little support for the hypothesis that the response of reservoir species to precipitation mediated the impact of precipitation on plague intensity. We instead observed that precipitation variables were of limited importance in defining species niches and rarely showed the expected response to precipitation across northern and southern China. These findings do not suggest that precipitation–reservoir species dynamics never influence plague intensity but that instead, the response of reservoir species to precipitation across a single biome cannot be assumed and that limited numbers of reservoir species may have a disproportional impact upon plague intensity.

Aim: Reconstructing species' glacial refugial history and demographic changes over time has greatly relied on comparing inferences from multiple methods while not sufficiently acknowledging their limitations. Here, we aim to integrate as fully as possible complementary methods in ecology, genomics and palaeobiology to improve the reconstruction of species biogeographical history. Location: South‐western Europe. Taxon: Cabrera vole (Microtus cabrerae). Methods: We compiled and mapped the raw fossil data of the Cabrera vole for the Last Glacial Maximum (LGM) and Mid‐Holocene (MH). Alongside, we projected the calibrated ecological niche model (ENM) of the species' current distribution for the LGM, the Younger Dryas and the MH. Complementarily, we used previously obtained Genotyping‐by‐Sequencing data to evaluate the demographic history and range expansion patterns of all four Evolutionarily Significant Units of the species, in an integrative framework. Results: ENM‐inferred refugial areas and genomic modelling consistently supported northern Iberian glacial refugia for the Cabrera vole. This contrasted with the higher fossil abundance of the species in southern and eastern Iberia and southern France from the LGM to the MH. Our results suggest that populations in areas with high fossil abundance went extinct, and were later replaced by northern Iberian populations such that they did not contribute significantly to the current gene pool. Main conclusions Our integrative approach indicates how the range of the Cabrera vole fluctuated in response to environmental change during and following the LGM. Despite methodological limitations, the ENM and genomic approaches produced generally congruent results. Instead, the fossil record may misrepresent the ancestral distribution of this species and should be considered cautiously for ancestral distribution reconstruction, considering that it also reflects the fossilization conditions. Overall, our study supports the idea that integrative approaches are essential to provide an accurate and well‐supported picture of historical refugial areas and range dynamics. [ABSTRACT FROM AUTHOR]

This data package enables empirical analysis of national electricity system transitions and retrospective evaluation of electricity system models in 1990–2019 in 31 European countries, including the EU27, Switzerland, Iceland, Norway, and the United Kingdom. The data package covers two types of content. Firstly, we provide an annotated list of 528 original data sources and references relevant for retrospective electricity system modeling with emphasis on open-access sources. Secondly, we provide 1359 processed data files in a format that is suitable as input to electricity system models. Four types of data files are included for each country: (i) a country file documenting national electricity demand and economic data, (ii) technology files describing techno-economic data for each major generation technology in the country's electricity mix, (iii) resource files describing prices and CO2 emissions for each generation fuel or input resource, and (iv) load profiles describing 24 h national load curves for each available year. We provide these data files as comma-separated files to enable their wider reuse for retrospective evaluation of models as well as for empirical analyses of the European electricity system transitions.

Long records of wave parameters are central to the estimation of coastal flooding risk and the causes of coastal erosion. This paper leverages the predictive power of wave height history and correlations with wind speed and direction to build statistical models for time series of wave heights to develop a method to fill data-gaps and extend the record length of coastal wave observations. A threshold regression model is built where the threshold parameter, based on lagged wind speed, explains the nonlinear associations, and the lagged predictors in the model are based on a well-established empirical wind-wave relationship. The predictive model is completed by addressing the residual conditional heteroscedasticity using a GARCH model. This comprehensive model is trained on time series data from 2005 to 2013, using wave height and wind data both observed from a buoy in Long Island Sound. Subsequently, replacing wind data with observations from a nearby coastal station provides a similar level of predictive accuracy. This approach can be used to hindcast wave heights for past decades given only wind information at a coastal station. These hindcasts are used as a representative of the unobserved past to carry out extreme value analysis by fitting Generalized Pareto (GP) distribution in a peaks over threshold (POT) framework. By analyzing longer periods of data, we can obtain reliable return value estimates to help design better coastal protection structures.

The reconstruction or prediction of meteorological records through the Analog Ensemble (AnEn) method is very efficient when the number of predictor time series is small. Thus, in order to take advantage of the richness and diversity of information contained in a large number of predictors, it is necessary to reduce their dimensions. This study presents methods to accomplish such reduction, allowing the use of a high number of predictor variables. In particular, the techniques of Principal Component Analysis (PCA) and Partial Least Squares (PLS) are used to reduce the dimension of the predictor dataset without loss of essential information. The combination of the AnEn and PLS techniques results in a very efficient hybrid method (PLSAnEn) for reconstructing or forecasting unstable meteorological variables, such as wind speed. This hybrid method is computationally demanding but its performance can be improved via parallelization or the introduction of variants in which all possible analogs are previously clustered. The multivariate linear regression methods used on the new variables resulting from the PCA or PLS techniques also proved to be efficient, especially for the prediction of meteorological variables without local oscillations, such as the pressure.

The Analogue Ensemble (AnEn) method enables the reconstruction of meteorological observations or deterministic predictions for a certain variable and station by using data from the same station or from other nearby stations. However, depending on the dimension and granularity of the historical datasets used for the reconstruction, this method may be computationally very demanding even if parallelization is used. In this work, the classical AnEn method is modified so that analogues are determined using K-means clustering. The proposed combined approach allows the use of several predictors in a dependent or independent way. As a result of the flexibility and adaptability of this new approach, it is necessary to define several parameters and algorithmic options. The effects of the critical parameters and main options were tested on a large dataset from real-world meteorological stations. The results show that adequate monitoring and tuning of the new method allows for a considerable improvement of the computational performance of the reconstruction task while keeping the accuracy of the results. Compared to the classical AnEn method, the proposed variant is at least 15-times faster when processing is serial. Both approaches benefit from parallel processing, with the K-means variant also being always faster than the classic method under that execution regime (albeit its performance advantage diminishes as more CPU threads are used). [ABSTRACT FROM AUTHOR]

Climate change and harvesting can affect the ecosystems' functioning by altering the population dynamics and interactions among species. Knowing how species interact is essential for better understanding potentially unintended consequences of harvest on multiple species in ecosystems. I analyzed how stage‐specific interactions between two harvested competitors, the haddock (Melanogrammus aeglefinus) and Atlantic cod (Gadus morhua), living in the Barents Sea affect the outcome of changes in the harvest of the two species. Using state‐space models that account for observation errors and stochasticity in the population dynamics, I run different harvesting scenarios and track population‐level responses of both species. The increasing temperature elevated the number of larvae of haddock but did not significantly influence the older age‐classes. The nature of the interactions between both species shifted from predator‐prey to competition around age‐2 to ‐3. Increased cod fishing mortality, which led to decreasing abundance of cod, was associated with an increasing overall abundance of haddock, which suggests compensatory dynamics of both species. From a stage‐specific approach, I show that a change in the abundance in one species may propagate to other species, threatening the exploited species' recovery. Thus, this study demonstrates that considering interactions among life history stages of harvested species is essential to enhance species' co‐existence in harvested ecosystems. The approach developed in this study steps forward the analyses of effects of harvest and climate in multi‐species systems by considering the comprehension of complex ecological processes to facilitate the sustainable use of natural resources. [ABSTRACT FROM AUTHOR]

The 22-year (1998–2019) surface seawater dimethylsulfide (DMS) concentrations in the Yellow Sea (YS) were hindcasted based on satellite sea surface temperature (SST) and chlorophyll-a (Chl-a) data using a generalized additive mixed model (GAMM). A continuous monthly dataset of DMS concentration in the YS was obtained after using the data interpolation empirical orthogonal function (DINEOF) to reconstruct missing information in the dataset. Then, the interannual DMS variability in the YS was analyzed. The results indicated that the monthly climatological DMS concentration in the YS was 3.61 nmol/L. DMS concentrations in the spring and summer were significantly higher than those in the autumn and winter. DMS concentrations were highest in coastal YS waters and lowest primarily in offshore YS waters. Interannual DMS variability between 1998 and 2019 was subdivided into two inverse phases: with the exception of the central YS, DMS increased before the turning point and decreased after. The turning point in interannual DMS variation was earlier in the inshore YS as compared to the central YS. Spectrum analysis identified some significant patterns of interannual variation in the DMS anomaly in the YS. Chl a appeared to be the main factor influencing interannual trends in DMS in the YS. Interannual DMS variability was under the joint control of Chl a and SST. However, short-term interannual DMS variation (2–3 years) was primarily related to SST, while longer term interannual DMS variation (6–8 years) was significantly correlated with Chl a and SST. [ABSTRACT FROM AUTHOR]

The western massasauga (Sistrurus tergeminus) is a small pit viper with an extensive geographic range, yet observations of this species are relatively rare. They persist in patchy and isolated populations, threatened by habitat destruction and fragmentation, mortality from vehicle collisions, and deliberate extermination. Changing climates may pose an additional stressor on the survival of isolated populations. Here, we evaluate historic, modern, and future geographic projections of suitable climate for S. tergeminus to outline shifts in their potential geographic distribution and inform current and future management. We used maximum entropy modeling to build multiple models of the potential geographic distribution of S. tergeminus. We evaluated the influence of five key decisions made during the modeling process on the resulting geographic projections of the potential distribution, allowing us to identify areas of model robustness and uncertainty. We evaluated models with the area under the receiver operating curve and true skill statistic. We retained 16 models to project both in the past and future multiple general circulation models. At the last glacial maximum, the potential geographic distribution associated with S. tergeminus occurrences had a stronghold in the southern part of its current range and extended further south into Mexico, but by the mid‐Holocene, its modeled potential distribution was similar to its present‐day potential distribution. Under future model projections, the potential distribution of S. tergeminus moves north, with the strongest northward trends predicted under a climate scenario increase of 8.5 W/m2. Some southern populations of S. tergeminus have likely already been extirpated and will continue to be threatened by shifting availability of suitable climate, as they are already under threat from desertification of grasslands. Land use and habitat loss at the northern edge of the species range are likely to make it challenging for this species to track suitable climates northward over time. [ABSTRACT FROM AUTHOR]

Abstract The western massasauga (Sistrurus tergeminus) is a small pit viper with an extensive geographic range, yet observations of this species are relatively rare. They persist in patchy and isolated populations, threatened by habitat destruction and fragmentation, mortality from vehicle collisions, and deliberate extermination. Changing climates may pose an additional stressor on the survival of isolated populations. Here, we evaluate historic, modern, and future geographic projections of suitable climate for S. tergeminus to outline shifts in their potential geographic distribution and inform current and future management. We used maximum entropy modeling to build multiple models of the potential geographic distribution of S. tergeminus. We evaluated the influence of five key decisions made during the modeling process on the resulting geographic projections of the potential distribution, allowing us to identify areas of model robustness and uncertainty. We evaluated models with the area under the receiver operating curve and true skill statistic. We retained 16 models to project both in the past and future multiple general circulation models. At the last glacial maximum, the potential geographic distribution associated with S. tergeminus occurrences had a stronghold in the southern part of its current range and extended further south into Mexico, but by the mid‐Holocene, its modeled potential distribution was similar to its present‐day potential distribution. Under future model projections, the potential distribution of S. tergeminus moves north, with the strongest northward trends predicted under a climate scenario increase of 8.5 W/m2. Some southern populations of S. tergeminus have likely already been extirpated and will continue to be threatened by shifting availability of suitable climate, as they are already under threat from desertification of grasslands. Land use and habitat loss at the northern edge of the species range are likely to make it challenging for this species to track suitable climates northward over time.

Migration allows animals to exploit spatially separated and seasonally available resources at a continental to global scale. However, responding to global climatic changes might prove challenging, especially for long-distance intercontinental migrants. During glacial periods, when conditions became too harsh for breeding in the north, avian migrants have been hypothesized to retract their distribution to reside within small refugial areas. Here, we present data showing that an Afro-Palearctic migrant continued seasonal migration, largely within Africa, during previous glacial-interglacial cycles with no obvious impact on population size. Using individual migratory track data to hindcast monthly bioclimatic habitat availability maps through the last 120,000 y, we show altered seasonal use of suitable areas through time. Independently derived effective population sizes indicate a growing population through the last 40,000 y. We conclude that the migratory lifestyle enabled adaptation to shifting climate conditions. This indicates that populations of resource-tracking, longdistance migratory species could expand successfully during warming periods in the past, which could also be the case under future climate scenarios. [ABSTRACT FROM AUTHOR]

Hindcasting of past levee failures enhances insights in the performance and vulnerability of levees. The scarcity of field evidence makes identifying the cause(s) of failure difficult. Under these circumstances, multiple scenarios and model choices are possible to characterise and to model the failure. This paper shows how probabilistic Bayesian techniques advance the procedure of hindcasting of levee failures. In the developed approach, a-priori levee information, and failure observations are systematically taken into account to determine the most likely scenario and the most representative model choices to characterise the failure most accurately. Observations, such as the slip surface, are taken into account in the probability estimates. The levee failure near Breitenhagen, Germany (2013) is used as a case study. The levee failed during river floods due the instability of the landside slope. The levee failure was most likely triggered by locally weak soil conditions and unexpected high water pressures due a connection between a pond on the riverside of the levee and the aquifer. These conditions were likely caused by the occurrence of a previous breach at this location. The approach developed in this paper is expected to support a more systematic and objective method of analysis of other levee failures. [ABSTRACT FROM AUTHOR]

Context: Environmental changes produce discontinuities in suitable habitat. However, drawing inference into the effects of these changes on contemporary genetic patterns is often difficult. Recent approaches for evaluating landscape resistance facilitate increased understanding of landscape effects on gene flow. Objectives: We investigated the effects of pluvial lakes and sagebrush cover on genetic connectivity for the pygmy rabbit (Brachylagus idahoensis), a sagebrush obligate species. We predicted that sagebrush-based surfaces would be more explanatory than pluvial lake surfaces. Furthermore, we predicted that habitat characteristics during the mid-Holocene would explain genetic differentiation better than those during the late-Pleistocene. Methods: We leveraged a genetic dataset to evaluate the explanatory power of landscape resistance surfaces. We generated resistance surfaces that represent varying degrees of resistance associated with pluvial lakes and sagebrush cover, then projected sagebrush distribution back to the mid-Holocene and late-Pleistocene. Results: Representations based on sagebrush distribution were more explanatory than those based on pluvial lakes. Projections of sagebrush distribution back in time indicate concordance between genetic connectivity and mid-Holocene sagebrush distribution. Limited numbers of dispersal pathways were apparent among study regions, suggesting spatially restricted corridors of connectivity. Conclusions: We demonstrate that shifts in vegetative cover can shape contemporary patterns of genetic connectivity. By coupling testing of resistance surfaces with estimates of past vegetative change, we provide insights into the time scales over which genetic differentiation may occur. Given projections of future declines in sagebrush, maintaining sagebrush cover will be critical to population persistence of pygmy rabbits. [ABSTRACT FROM AUTHOR]

Globally, anthropogenic land‐cover change has been dramatic over the last few centuries and is frequently invoked as a major cause of wildlife population declines. Baseline data currently used to assess population trends, however, began well after major changes to the landscape. In the United States and Canada, breeding bird population trends are assessed by the North American Breeding Bird Survey, which began in the 1960s. Estimates of distribution and abundance prior to major habitat alteration would add historical perspective to contemporary trends and allow for historically based conservation targets. We used a hindcasting framework to estimate change in distribution and abundance of 7 bird species in the Willamette Valley, Oregon (United States). After reconciling classification schemes of current and 1850s reconstructed land cover, we used multiscale species distribution models and hierarchical distance sampling models to predict spatially explicit densities in the modern and historical landscapes. We estimated that since the 1850s, White‐breasted Nuthatch (Sitta carolinensis) and Western Meadowlark (Sturnella neglecta) populations, 2 species sensitive to fragmentation of oak woodlands and grasslands, declined by 93% and 97%, respectively. Five other species we estimated nearly stable or increasing populations, despite steep regional declines since the 1960s. Based on these estimates, we developed historically based conservation targets for amount of habitat, population, and density for each species. Hindcasted reconstructions provide historical perspective for assessing contemporary trends and allow for historically based conservation targets that can inform current management. Article impact statement: Hindcasted estimates of historical wildlife populations can inform current conservation targets. [ABSTRACT FROM AUTHOR]

The Analogue Ensemble (AnEn) method enables the reconstruction of meteorological observations or deterministic predictions for a certain variable and station by using data from the same station or from other nearby stations. However, depending on the dimension and granularity of the historical datasets used for the reconstruction, this method may be computationally very demanding even if parallelization is used. In this work, the classical AnEn method is modified so that analogues are determined using K-means clustering. The proposed combined approach allows the use of several predictors in a dependent or independent way. As a result of the flexibility and adaptability of this new approach, it is necessary to define several parameters and algorithmic options. The effects of the critical parameters and main options were tested on a large dataset from real-world meteorological stations. The results show that adequate monitoring and tuning of the new method allows for a considerable improvement of the computational performance of the reconstruction task while keeping the accuracy of the results. Compared to the classical AnEn method, the proposed variant is at least 15-times faster when processing is serial. Both approaches benefit from parallel processing, with the K-means variant also being always faster than the classic method under that execution regime (albeit its performance advantage diminishes as more CPU threads are used).

This study concerns the problem of making weather predictions for a location where no data is available, using meteorological datasets from nearby stations. The hindcast with multiple stations is performed with different variants of the Analog Ensemble (AnEn) method. In addition to the traditional Monache metric used to identify analogs in datasets from one or two stations, several new metrics are explored, namely cosine similarity, normalization and k-means clustering. These were analyzed and benchmarked to find the ones that bring improvements. The best results were obtained with the k-means metric, yielding between 3% and 30% of lower quadratic error when compared against the Monache metric. Also, making the predictors include two stations improved the performance of the hindcast, leading up to 16% of lower error, depending on the correlation between the predictor stations.

A comparative study between classic linear and intelligent nonlinear time series approaches for short-term maximum wave height forecasting is presented in this study. The applied models to accomplish a use case for onshore measurements from the Mediterranean Sea include ordinary linear regression (LR), autoregressive integrated moving average (ARIMA), artificial neural networks (ANN), and genetic programming (GP). The study also introduces a new evolutionary ensemble model called ensemble GP, which integrates effective models' forecasts through an evolutionary procedure. The results from standalone models showed that both linear and nonlinear models provide the same accuracy for short-term maximum wave height hindcasting on a seasonal scale. The proposed ensemble model can enhance the forecasting accuracy of standalone models markedly. The new model can forecast maximum wave heights with the root mean squared errors less than 5 cm and Nash-Sutcliff efficiency more than 0.97. It is explicit and secures parsimony conditions, thus it is proposed to be used in practice. [ABSTRACT FROM AUTHOR]

Lake Ontario ice conditions are statistically linked to regional temperatures recorded in Toronto, during the most recent climate normal (1980/81–2009/10). A metric was developed to capture the net melting effect of average winter temperatures to characterize lake ice conditions, referred to as Net Melting-Degree Days (NMDD). This metric was able to account for 78% of lake ice interannual variability (R2 = 0.783, P < 0.001). Based on NMDD parameters, current lake ice conditions were characterized in four ways: heavy, moderate, light and very light. Lake Ontario ice conditions were reconstructed to create a hindcast for the span of the instrumental temperature record (1840/41–1979/80). Based on a decadal analysis, heavy ice seasons decreased significantly (R2 = 0.658, P < 0.001) from the 1840s to the 2000s, declining from an average of 6 heavy ice seasons per decade during the most distant climate normal (1840s to 1960s) to an average of only 1 heavy ice season per decade during the most recent climate normal (1980s to 2000s). Finally, lake ice conditions are projected to the end of the 21st century, using an optimal ensemble of Global Climate Model outputs for two different climate change scenarios (RCP4.5, RCP8.5). Heavy ice seasons no longer occur as early as the 2050s under both RCP4.5 and RCP8.5. Whereas, very light ice seasons go from being an extreme in the baseline period (10%), to the dominant characterization of Lake Ontario ice conditions by the 2080s, for both RCP4.5 (73%) and RCP8.5 (100%). [ABSTRACT FROM AUTHOR]

The amount and extent of coastal flooding caused by hurricanes can be sensitive to the timing or speed of the storm. For storms moving parallel to the coast, the hazards can be stretched over a larger area. Hurricane Matthew was a powerful storm that impacted the southeastern U.S. during October 2016, moving mostly parallel to the coastline from Florida through North Carolina. In this study, three sources for atmospheric forcing are considered for a simulation of Matthew's water levels, which are validated against extensive observations, and then the storm's effects are explored on this long coastline. It is hypothesized that the spatial variability of Matthew's effects on total water levels is partly due to the surge interacting nonlinearly with tides. By changing the time of occurrence of the storm, differences in storm surge are observed in different regions due to the storm coinciding with other periods in the tidal cycles. These differences are found to be as large as 1 m and comparable to the tidal amplitude. A change in forward speed of the storm also should alter its associated flooding due to differences in the duration over which the storm impacts the coastal waters. With respect to the forward speed, the present study contributes to established results by considering the scenario of a shore-parallel hurricane. A faster storm caused an increase in peak water levels along the coast but a decrease in the overall volume of inundation. On the other hand, a slower storm pushed more water into the estuaries and bays and flooded a larger section of the coast. Implications for short-term forecasting and long-term design studies for storms moving parallel to long coastlines are discussed herein. • OWI winds represent Hurricane Matthew's effects more accurately than WF and GAHM. • Nonlinear interaction between surge and tide can alter storm tide significantly. • A shore-parallel storm can interact with various stages in the tidal cycle. • A faster storm would have caused higher surges on the open coast. • A slower storm would have pushed more water into the estuaries and bays. [ABSTRACT FROM AUTHOR]

The Soil Moisture Active Passive (SMAP) mission measures important soil moisture data globally. SMAP’s products might not always perform better than land surface models (LSM) when evaluated against in situ measurements. However, we hypothesize that SMAP presents added value for long-term soil moisture estimation in a data fusion setting as evaluated by in situ data. Here, with the help of a time series deep learning (DL) method, we created a seamlessly extended SMAP data set to test this hypothesis and, importantly, gauge whether such benefits extend to years beyond SMAP’s limited lifespan. We first show that the DL model, called long short-term memory (LSTM), can extrapolate SMAP for several years and the results are similar to the training period. We obtained prolongation results with low-performance degradation where SMAP itself matches well with in situ data. Interannual trends of root-zone soil moisture are surprisingly well captured by LSTM. In some cases, LSTM’s performance is limited by SMAP, whose main issue appears to be its shallow sensing depth. Despite this limitation, a simple average between LSTM and an LSM Noah frequently outperforms Noah alone. Moreover, Noah combined with LSTM is more skillful than when it is combined with another LSM. Over sparsely instrumented sites, the Noah–LSTM combination shows a stronger edge. Our results verified the value of LSTM-extended SMAP data. Moreover, DL is completely data driven and does not require structural assumptions. As such, it has its unique potential for long-term projections and may be applied synergistically with other model-data integration techniques. [ABSTRACT FROM AUTHOR]

The ability to forecast air temperature and precipitation at monthly-seasonal time scale in the Amur River basin is assessed. The climatic characteristics of actual monthly values of air temperature and total precipitation are compared with those based on the semi-Lagrangian global finite-difference model (SL-AV) of general atmospheric circulation for the Amur basin. It is shown that there is high correlation between total precipitation for 1–2 months and water discharge in the Amur River in next months during the warm half-year. The SL-AV model forecasts of air temperature and precipitation in the Amur basin are verified. It is demonstrated that the accuracy of prediction of total precipitation averaged over the Amur basin is higher as compared with that for initial gridded forecasts. The accuracy of such forecasts is rather high for July and August being critical months for the flood formation. The procedure for the statistical correction of monthly and seasonal forecasts of air temperature and precipitation based on the SL-AV model and the regression scheme for the Amur runoff prediction are proposed. [ABSTRACT FROM AUTHOR]

Vegetation dynamics play a crucial role in understanding the complex interactions between climate variability, ecosystem changes and human land use. In this study, we use the Hindcasting Ecosystems Model (HEMO) to reconstruct the paleo Normalized Difference Vegetation Index (paleo-NDVI), which provides valuable insights into the historical vegetation patterns. Our research focuses on the El Alto Ancasti mountain range, a region characterized by its rich environmental diversity, and archaeological evidence of pre-Hispanic agricultural village landscapes, showcasing the knowledge and management of diverse environments and their fluctuations. To reconstruct the paleo-NDVI, we employed a Neuronal Network trained with two datasets. First, we utilized NDVI data extracted from AVHRR/NOAA satellite images spanning from 1982 to 2015, representing 8 km pixels from the Global Inventory Modeling and Mapping Studies. Second, we incorporated growth ring chronologies from Northwest Argentina sourced from the International Tree Ring Data Bank, enabling the calibration of the NDVI-ring relationship. This methodology allowed us to reconstruct the paleo-NDVI back to 442 AD and to calculate anomalies. Our analysis revealed significant negative anomalies around 600 AD, coinciding with previous research conducted in the nearby Ambato valley. These anomalies may be attributed to the eruption of Tres Cruces volcano in Catamarca during that period. Additionally, extensive positive anomalies emerged from 1600 AD onwards, which align with anomalies detected in earlier studies. The eruption of Huaynaputina volcano in southern Peru in 1600 AD is a possible factor contributing to these positive anomalies. Ongoing investigations focusing on the paleoenvironment and human occupations in the El Alto-Ancasti area aim to provide further insights into how these detected changes could have influenced environmental management decisions. By integrating multidisciplinary approaches, we can better comprehend the complex relationship between vegetation dynamics, climatic events and human activities, ultimately enhancing our understanding of past environmental transformations and their implications for sustainable management strategies. • Use of tree rings and NDVI data for paleoenvironmental retrodictions by means of artificial neural networks (NN). • Retrodiction of paleo-NDVI up to 442DC in Alto Ancasti (Catamarca, Argentina) by means of artificial neural networks (NN). • Coherence between paleo-NDVI retrodiction of Sierra Alto Ancasti with that of the Ambato Valley. • Finding of relation between paleo-NDVI variations and climate forcings such as volcanic events and ENSO. • Evidence of interrelations between ecosystem dynamics and environmental management of past occupations. [ABSTRACT FROM AUTHOR]

The success of a strategy of mitigation of the effects of the droughts requires the implementation of an effective monitoring and forecasting system, able to identify drought events and follow their spatiotemporal evolution. This article demonstrates the capability of the artificial neural networks in predicting the spring standardized precipitation index, SPI, for Portugal. The validation of the models used the hindcasting, which is a technique by which a given model is tested through its application to historical data followed by the comparison of the results thus achieved with the data. The SPI index was calculated at the timescale of six months and the climate indices used as external predictors in the hindcasting were the North Atlantic Oscillation and temperatures of the sea surface. The study showed the added value of the inclusion of previous predictors in the model. Maps of the probabilities of the drought occurrences which may be very important for integrated planning and management of water resources were also developed.

Abstract: Habitat loss and degradation induced by human development are among the major threats to biodiversity worldwide. In this study, we tested our ability to predict the response of bird communities (128 species) to land‐use changes in southern Quebec (~483,100 km2) over the last 30 yr (between 1984–1989 and 2010–2014) by using species distribution models (299,302 occurrences in 30,408 locations) from a hindcasting perspective. Results were grouped by functional guilds to infer potential impacts on ecosystem services, and to relate model transferability (i.e., ability of our models to be generalized to other times and scales) to specific functional and life‐history traits. Overall, our models were able to accurately predict, both in space and time, habitat suitability for 69% of species, especially for granivorous, nonmigrant, tree‐nesting species, and species that are tied to agricultural areas under intensive use. These findings indicate that model transferability depends upon specific functional and life‐history traits, providing further evidence that species’ ecologies affect the ability of models to accurately predict bird distributions. Declining bird species were mostly short‐distance migrants that were associated with open habitats (agricultural and nonproductive forest) with aerial insectivorous or granivorous diets, which may be related to agricultural intensification and land abandonment. Land‐use changes were positive for some forest bird species that were mainly associated with mixed and deciduous forests, generalist diets and tree‐nesting strategies. Yet cavity‐nesting birds have suffered substantial reductions in their distributions, suggesting that cumulative effects of intensive logging and wildfires on mature forests pose a threat for forest‐specialist species. Habitat suitability changes predicted by our coarse‐scale species distribution models partially agreed with the long‐term trends reported by the North American Breeding Bird Survey. Our findings confirm land‐use change as a key driving force for shaping bird communities in southern Quebec, together with the need to explicitly incorporate it into global change scenarios that better inform decision‐makers on conservation and management. [ABSTRACT FROM AUTHOR]

Abstract: Aim: Range shift is a relatively well‐understood response to climate change, but our ability to predict shifts is limited. Two factors that may cause variation in range shifts across species are dispersal ability and varying rates of climate change through time and across space. Here, we assess patterns of range shifts during the late Quaternary and estimate how the velocity of climate change and the dispersal ability of a species affect the magnitude of species range shifts in response to climate change. Location: North America. Methods: We hindcast species distribution models for 122 North American mammals to five times over the past 17,000 years and forecast them to two future times given two emissions scenarios. Generalized additive models were constructed to quantify the importance of dispersal ability and the velocity of temperature and precipitation in determining the magnitude of range shift expected for individual species. Results: Hindcasted and forecasted ranges demonstrate the variety of responses to climate change. In general, species shifted their ranges in a northerly direction (NW, N, NE) regardless of the type of climate change (i.e., warming vs. cooling). The highest rates of range shifts during the past occurred during periods of relatively rapid climate change (Last Glacial Maximum/Bølling‐Allerød and Bølling‐Allerød/Younger Dryas transitions). Rates of range shifts for the future are projected to be significantly higher than any of the past intervals. The velocity of climate change is significantly associated with the magnitude of range shifts during climate transitions that occur over longer time‐scales, while maximum dispersal distance is important during periods of rapid climate change. Main Conclusions: Our results suggest that both the dispersal ability and the velocity of climate change are significantly associated with species’ range shifts. However, the importance of these two factors is context‐dependent and depends on the interaction of the rate of climate change and the length of time over which the change occurs. [ABSTRACT FROM AUTHOR]

SWAN (Simulating Wave Near-shore) is a numerical wave model for hindcasting/forecasting wave parameters in coastal areas. This numerical model is chosen because is suitable for shallow water. This study was conducted to verify the results of wave height hindcasting in Jepara coastal waters. This is expected to support wave characteristic research based on wave forecasting for 10 years in the waters between Java, Sumatera and Kalimantan. The model is run with the third-generation mode (GEN3), which allow wind input, quadruplet and triad interactions, whitecapping, and breaking. Wind data is obtained from ECMWF (European Centre for Medium-Range Weather Forecasts) and the bathymetry from GEBCO (General Bathymetric Chart of The Oceans). The validation of the model and buoy data during July - December 1993 shows a good result (Root Mean Square Error = 0.166 and correlation/ linear regression = 0.807). Based on the literature, qualitatively the model has been verified with other simulation from another model in the same location. [ABSTRACT FROM AUTHOR]

Cannibalism is known to be a significant source of natural mortality of young North-east Arctic (NEA) cod. Cannibalism data, starting from 1984, have been used in NEA cod stock assessments since 1995, which has led to inconsistency in the cod abundance time series from 1946 to the present. To address this inconsistency, this study estimates the cannibalism-induced mortality (M2) of NEA cod at age 3–5 for the period 1946–1983. Combined qualitative and quantitative cod stomach content data for 1984–2010 were used to make the M2 estimations for age groups 3–5 (ICES 2014), then different factors including SSB were used to examine which covariates explained variability in M2 and thus make predictions for 1946–1983. The level of cannibalism was estimated to be high in the 1950s – early1960s. VPA-based assessment was run using these estimated M2 values. As a result, numbers of cod eaten by their conspecifics in the historical period and new increased recruitment estimates at age 3 were computed. The main factors affecting cannibalism appeared to be young cod abundance, total stock biomass (TSB) of large cod, and capelin total stock biomass (which represents an alternative prey). The problems involved in using the new recruitment time series in fishery management are discussed. The methodology presented here represents a generic approach to extending predation mortalities back in time to improve historical stock estimates. [ABSTRACT FROM PUBLISHER]

Over the last few decades, several efforts have been made to develop an alternative and sustainable energy source from wind waves. To achieve financial sustainability of this technology, most of the research has focused on analyzing facilities composed of several wave energy converters (WECs) arrays instead of isolated ones. Although the interaction between devices and its implications on the performance of the facilities have been studied previously, these works considered only certain combinations of sea states, limiting the applicability of the results. This work applies a new methodology based on statistical methods to assess the performance of different WEC array distributions during their entire life-cycle in an efficient way, using downscaling techniques and advanced numerical modeling to propagate the wave climate. The results obtained during the hindcasting life-cycle are used to analyze the maintenance and operation capabilities of the different alternatives of arrays defined for the WEC facility. The interactions between devices and their efficiency considering the associated impact are also quantified. The assessment of these efficiencies during the complete life-cycle of the devices is highly valuable tool for promoters and coastal managers to evaluate different WEC array alternatives. The entire process was applied to a hypothetical array location in the Gulf of Cádiz (southwestern Spain), where three different array distributions were defined. The results show that the distance between WECs is a key parameter that controls the potential energy production, the efficiency of the facility and the interactions between several devices. [ABSTRACT FROM AUTHOR]

The Soil Moisture Active Passive (SMAP) mission has delivered valuable sensing of surface soil moisture since 2015. However, it has a short time span and irregular revisit schedules. Utilizing a state-of-the-art time series deep learning neural network, Long Short-Term Memory (LSTM), we created a system that predicts SMAP level-3 moisture product with atmospheric forcings, model-simulated moisture, and static physiographic attributes as inputs. The system removes most of the bias with model simulations and improves predicted moisture climatology, achieving small test root-mean-square errors (<0.035) and high-correlation coefficients >0.87 for over 75% of Continental United States, including the forested southeast. As the first application of LSTM in hydrology, we show the proposed network avoids overfitting and is robust for both temporal and spatial extrapolation tests. LSTM generalizes well across regions with distinct climates and environmental settings. With high fidelity to SMAP, LSTM shows great potential for hindcasting, data assimilation, and weather forecasting. [ABSTRACT FROM AUTHOR]

In order to improve the forecasting ability of numerical models, a sequential data assimilation scheme, nudging, was applied to blend remotely sensing high-frequency (HF) radar surface currents with results from a three-dimensional numerical, EFDC (Environmental Fluid Dynamics Code) model. For the first time, this research presents the most appropriate nudging parameters, which were determined from sensitivity experiments. To examine the influence of data assimilation cycle lengths on forecasts and to extend forecasting improvements, the duration of data assimilation cycles was studied through assimilating linearly interpolated temporal radar data. Data assimilation nudging parameters have not been previously analyzed. Assimilation of HF radar measurements at each model computational timestep outperformed those assimilation models using longer data assimilation cycle lengths; root-mean-square error (RMSE) values of both surface velocity components during a 12 h model forecasting period indicated that surface flow fields were significantly improved when implementing nudging assimilation at each model computational timestep. The Data Assimilation Skill Score (DASS) technique was used to quantitatively evaluate forecast improvements. The averaged values of DASS over the data assimilation domain were 26% and 33% for east-west and north-south velocity components, respectively, over the half-day forecasting period. Correlation of Averaged Kinetic Energy (AKE) was improved by more than 10% in the best data assimilation model. Time series of velocity components and surface flow fields were presented to illustrate the improvement resulting from data assimilation application over time. [ABSTRACT FROM AUTHOR]

Understanding the effects of past climatic fluctuations on the distribution and population-size dynamics of cold-adapted species is essential for predicting their responses to ongoing global climate change. In spite of the heterogeneity of cold-adapted species, two main contrasting hypotheses have been proposed to explain their responses to Late Quaternary glacial cycles, namely, the interglacial contraction versus the interglacial expansion hypotheses. Here, we use the cold-adapted plant Primula farinosa to test two demographic models under each of the two alternative hypotheses and a fifth, null model. We first approximate the time and extent of demographic contractions and expansions during the Late Quaternary by projecting species distribution models across the last 72 ka. We also generate genome-wide sequence data using a Reduced Representation Library approach to reconstruct the spatial structure, genetic diversity, and phylogenetic relationships of lineages within P. farinosa. Finally, by integrating the results of climatic and genomic analyses in an Approximate Bayesian Computation framework, we propose the most likely model for the extent and direction of population-size changes in P. farinosa through the Late Quaternary. Our results support the interglacial expansion of P. farinosa, differing from the prevailing paradigm that the observed distribution of cold-adapted species currently fragmented in high altitude and latitude regions reflects the consequences of postglacial contraction processes. [ABSTRACT FROM AUTHOR]

We analyze how remote sounding instruments can help to improve our understanding of the atmospheric boundary layer and how regional synoptic models can be used as a hindcasting tool in studying and perfecting boundary layer models. A method is suggested for estimating the quality of boundary layer reproduction in these models using remote sensing data. [ABSTRACT FROM AUTHOR]

The increasing development of species distribution models (SDMs) using palaeodata has created new prospects to address questions of evolution, ecology and biogeography from wider perspectives. Palaeobotanical data provide information on the past distribution of taxa at a given time and place and its incorporation on modelling has contributed to advancing the SDM field. This has allowed, for example, to calibrate models under past climate conditions or to validate projected models calibrated on current species distributions. However, these data also bear certain shortcomings when used in SDMs that may hinder the resulting ecological outcomes and eventually lead to misleading conclusions. Palaeodata may not be equivalent to present data, but instead frequently exhibit limitations and biases regarding species representation, taxonomy and chronological control, and their inclusion in SDMs should be carefully assessed. The limitations of palaeobotanical data applied to SDM studies are infrequently discussed and often neglected in the modelling literature; thus, we argue for the more careful selection and control of these data. We encourage authors to use palaeobotanical data in their SDMs studies and for doing so, we propose some recommendations to improve the robustness, reliability and significance of palaeo-SDM analyses. [ABSTRACT FROM AUTHOR]

Long-term records of phytoplankton blooms in freshwater lakes are necessary both for understanding basin-scale changes to watersheds and for providing a key constraint for assessing processes driving blooms. However, due to the inherent constraints of in situ sampling and the short time period covered by modern space borne sensors, few long-term records exist. Historical data from sensors such as Landsat offer strong potential for creating new records of past blooms. Here, we use a novel evaluation procedure based on multiple metrics to assess algorithm suitability and robustness for generating long-term bloom records using Landsat 5 imagery. Evaluation metrics are based on bloom presence, spatial distribution, magnitude and timing, using both in situ Microcystis biovolume and remotely-sensed Cyanobacterial Index (CI) data from MERIS for 2002–2011. Applying this procedure for a test case focusing on Lake Erie's western basin, an algorithm based on a near infrared threshold with simple atmospheric correction through subtraction of the shortwave infrared band, combined with an additional “greenness” filter based on a hue threshold, performs best. Implementing this algorithm for 1984–2001 reveals the long-term trends in peak bloom magnitude prior to the start of the MERIS and MODIS record (2002–2015), and more than doubles the period of record that can be used to understand bloom occurrence and growth for this system. More broadly, we demonstrate that Landsat observations can be used to identify macro-scale features of blooms. For Lake Erie, the performance of the final Landsat algorithm is comparable to that of the MERIS CI algorithm, despite Landsat's broad spectral bands and long revisit time. [ABSTRACT FROM AUTHOR]

Aim Ecological niche models ( ENMs) are used widely in ecology, evolution, global change biology, but model uncertainty remains an underappreciated issue. Generally, either a single model from one algorithm or an ensemble of single models from different algorithms is used to provide a prediction. In addition to variability among algorithms, recent studies have shown the need to consider variability within a single algorithm, for example optimizing model complexity by tuning model settings. We present an ensemble ENM using a single-algorithm approach, while adjusting model settings to maximize performance. Location Madagascar. Methods We used maxent, bioclimatic variables and occurrence records of four species of Malagasy tenrecs (Family Tenrecidae). We calibrated and evaluated preliminary models using a jackknife approach, tuning two model settings to estimate optimal model complexity. We chose a suite of top-performing preliminary models and then generated a consensus prediction. Furthermore, we calculated the variability among predictions of the co-optimal models to indicate variation in geography (i.e. uncertainty). We then did the same after projecting the predictions to climatic estimates for the Last Glacial Maximum and the year 2070. Results The default settings were never identified as optimal for any of the four species. The model settings considered as the co-optimal solutions essentially led to the same evaluation statistics; however, they showed high variation in their geographic predictions for three of the four species. Additionally, variation among such models was greater when transferred across time. Main conclusions This approach likely can provide better predictions for a single algorithm as well as quantifications of within-algorithm uncertainty, qualities that are highly useful in interpreting reconstructed suitable areas or forecasts of potential range shifts under future climate change. Finally, this within-algorithm uncertainty can be integrated into a larger framework that considers variability due to other factors (e.g. related to input data, alternate algorithms or various Global circulation models). [ABSTRACT FROM AUTHOR]

The response of microalgae to environmental variation affects commercial scale production of algal feedstocks. Regionally-resolved predictive models of biomass productivity use climatic factors such as temperature and light to estimate algal growth and identify the best locations for production sites. However, such predictions have rarely been validated at continental scales due to the lack of continuous robust datasets originating from multiple locations simultaneously. We established and maintained cultures of Nannochloropsis sp. in open raceway ponds in 150 to 260 L for >17 months at three locations: Kailua-Kona, HI; Las Cruces, NM; and San Diego, CA. We compared predicted biomass productivities, based on local weather conditions, with experimental field data. We found considerable deviations between predicted and observed productivities across and within sites. Although weather remains a pertinent driver of productivity, site-specific factors, such as local pest pressures, chemical/biological contaminants, and differences in the surrounding environment heavily influence biomass productivities. Identifying and quantifying these local factors is key to formulating more accurate models to assess potential productivity and identify favorable sites for establishing algal production facilities. [Display omitted] • Robust dataset of simultaneous algal cultivation from multiple locations • Comparison of predicted biomass productivities with experimental field data • Algal productivity not explained solely by local climatic conditions. • Site-specific conditions heavily influence biomass productivities. [ABSTRACT FROM AUTHOR]

• Hindcasting to inform electricity system transition models in 31 European countries. • Testing of model versions with endogenous demand ED and technology learning TL. • ED functionality captures real-world transition with well estimated elasticity factor. • TL boosts earlier and larger uptake of renewable technologies in some countries. • TL can introduce inaccuracies as compared to the real-world transition. Bottom-up, technology-rich electricity system models are commonly used to generate scenarios for policy support at a national or global level. In literature, previous hindcasting studies (also called retrospective modeling or ex-post modeling) evaluated existing models rather than sought to inform model development from the beginning. In this study, we present a hindcasting exercise with D-EXPANSE model for national electricity systems in 31 European countries over the 1990–2019 period. We develop several model versions with or without elastic electricity demand and with or without endogenous technology learning, and use hindcasting to choose the most accurate configuration of the bottom-up model. The hindcasting results show that a model with endogenous elastic demand can capture well the real-world evolution of electricity demand, if elasticity factor is chosen well and if the countries did not undergo severe structural changes. Endogenous technology learning, however, increases the uptake of new emerging technologies in cost-optimal scenarios, but still cannot fully capture the real-world dynamics and at times even introduces further inaccuracies. [ABSTRACT FROM AUTHOR]

Species distribution models (SDMs) are a tool for predicting the eventual geographical range of an emerging pathogen. Most SDMs, however, rely on an assumption of equilibrium with the environment, which an emerging pathogen, by definition, has not reached. To determine if some SDM approaches work better than others for modelling the spread of emerging, non-equilibrium pathogens, we studied time-sensitive predictive performance of SDMs for Batrachochytrium dendrobatidis, a devastating infectious fungus of amphibians, using multiple methods trained on time-incremented subsets of the available data. We split our data into timeline-based training and testing sets, and evaluated models on each set using standard performance criteria, including AUC, kappa, false negative rate and the Boyce index. Of eight models examined, we found that boosted regression trees and random forests performed best, closely followed by MaxEnt. As expected, predictive performance generally improved with the length of time series used for model training. These results provide information on how quickly the potential extent of an emerging disease may be determined, and identify which modelling frameworks are likely to provide useful information during the early phases of pathogen expansion.

This article is concerned with attempting to ‘predict’ (hindcast) the damage caused by the L’Aquila 2009 earthquake (Mw 6.3) and, more generally, with the question of how close predicted damage can ever be to observations. Damage is hindcast using a well-established empirical-based approach based on vulnerability indices and macroseismic intensities, adjusted for local site effects. Using information that was available before the earthquake and assuming the same event characteristics as the L’Aquila mainshock, the overall damage is reasonably well predicted but there are considerable differences in the damage pattern. To understand the reasons for these differences, information that was only available after the event were include within the calculation. Despite some improvement in the predicted damage, in particularly by the modification of the vulnerability indices and the parameter influencing the width of the damage distribution, these hindcasts do not match all the details of the observations. This is because of local effects: both in terms of the ground shaking, which is only detectable by the installation of a much denser strong-motion network and a detailed microzonation, and in terms of the building vulnerability, which cannot be modeled using a statistical approach but would require detailed analytical modeling for which calibration data are likely to be lacking. Future studies should concentrate on adjusting the generic components of the approach to make them more applicable to their location of interest. To increase the number of observations available to make these adjustments, we encourage the collection of damage states (and not just habitability classes) following earthquakes and also the installation of dense strong-motion networks in built-up areas.

The paper presents a first experimental version of the original QGIS plugin QWaveTransposition that numerically implements the geographic transposition of wave gauge data method proposed by Contini and De Girolamo (1998) for offshore wave hindcasting. The method allows one to transfer wave data measured at a given gauging station to a virtual station located offshore the area of interest, by comparing the effective fetches at both stations. The QWaveTransposition plugin was implemented in Python programming language, including the NumPy package for numerical computations. A graphical user interface was developed to manage the input/output data and model parameters. The fetch geometry at real and virtual stations can be imported by selecting appropriate vector layers from the QGIS map. An application to a sample site in southern Italy is presented for example purposes. [ABSTRACT FROM AUTHOR]

This paper performs energy model hindcasting which compared the historical energy simulation results with the observations. We used one of the Integrated Assessment Models and simulated global historical energy consumption from 1981 to 2010 associated with exogenous socioeconomic assumptions, as is typically performed for future scenario. The simulation period was chosen with consideration of data availability and structural constancy of the model. Based on comparison with observations, there are three main findings. First, the global aggregated primary energy shows high reproducibility. In terms of energy source specific results, the fitness in electricity, coal, and biomass consumption were high. However, that of crude oil and natural gas is lower than others. This could be due to the price elasticity assumption, implying that the model can be improved with regard to this element. Second, the reproducibility increases as the simulation is close to the base year 2005. Third, although the global aggregated information shows high reproducibility, some disaggregated regions have lower reproducibility. Furthermore, high income countries tend to show higher reproducibility than in low income countries. Given the uncertainties in the ability of IAMs to reproduce certain aspects of the energy system, forecasts must be treated with caution. [ABSTRACT FROM AUTHOR]