Multi-model Approach Research Articles

Response of the American equatorial ionization anomaly to 2016 Arctic sudden stratospheric warming events

Apart from the rapid ionospheric response to geomagnetic forcing originating from the Sun during extreme space weather events, forcing from the lower atmosphere below still exerts a significant influence on the ionosphere during quiet-time conditions. This study examines the ionospheric response of the equatorial ionization anomaly (EIA) in the American sector to the combined influence of the cascades of sudden stratospheric warming (SSW) events and the geomagnetic storms that coexisted with them during the period of January–March 2016. We adopted a multi-instrument and multi-modeling approach with the study locations spanning ±40° geomagnetic latitudes. Our results showed a hemispheric asymmetry in the total electron content and change in total electron content (ΔTEC) distribution with higher enhancement clearly visible in the Northern Hemisphere in comparison to the Southern Hemisphere (NH). Semidiurnal signatures were observed in both ΔTEC and equatorial electrojet parameters for some days. The double-peak zonal mean zonal wind amplitude days supported the formation of the reverse fountain effects. The different SSW peak temperature days also showed either positive or negative ionospheric response. Generally, orientation of the prompt penetration electric field (PPEF) and their strengths at either daytime or nighttime played a weak role in the ionosphere response during some of the geomagnetic storms. The negative and positive ionospheric responses under geomagnetic storm conditions were ascribed to changes in the composition of the thermosphere, prompt penetration electric field (PPEF), and traveling atmospheric disturbances (TADs).

A Multi-Model Multi-Scale Approach to Estimate the Impact of the 2007 Large-Scale Forest Fires in Peloponnese, Greece

The hydrological impact of large-scale forest fires in a large basin is investigated on both a daily and an hourly basis. A basin of 877 km2 was chosen, with 37% of its area having been burnt in the summer of 2007. Five models are employed, namely SWAT (semi-distributed), GR4J, GR5J, and GR6J (lumped) for the daily time step, and HEC-HMS (semi-distributed) for the hourly time step. As SWAT and HEC-HMS implement the SCS-CN method, the change in the Curve Number (CN) from pre-fire to post-fire conditions is estimated along with the post-fire trend of CN for both time steps. Regarding the daily time step, a 20% post-fire increase in CN proved necessary for the accurate streamflow prediction, whereas ignoring this led to an underestimation of 22% on average. On an hourly time basis, CN was 95 for burnt areas after the fire, with a mildly decreasing trend after the third year and still above 90 until the fifth year. When neglecting this, peak flow is seriously underestimated (35–70%). The post-fire trend lines of CN for the two-time steps showed statistically equal slopes. Finally, GR models accurately predicted runoff while constraining one model parameter, which proved useful for the realistic prediction of other variables.

Length–Weight Relationships, Growth Models of Two Croakers (Pennahia macrocephalus and Atrobucca nibe) off Taiwan and Growth Performance Indices of Related Species

Information on age and growth is essential to modern stock assessment and the development of management plans for fish resources. To provide quality otolith-based estimates of growth parameters, this study performed five types of analyses on the two important croakers that were under high fishing pressure in southwestern Taiwan: Pennahia macrocephalus (big-head pennah croaker) and Atrobucca nibe (blackmouth croaker): (1) Estimation of length–weight relationships (LWR) with discussion on the differences with previous studies; (2) validation of the periodicity of ring formation using edge analysis; (3) examination of three age determination methods (integral, quartile and back-calculation methods) and selection of the most appropriate one using a k-fold cross-validation simulation; (4) determination of the representative growth models from four candidate models using a multimodel inference approach; and, (5) compilation of growth parameters for all Pennahia and Atrobucca species published globally for reviewing the clusters of estimates using auximetric plots of logged growth parameters. The study observed that features of samples affected the LWR estimates. Edge analysis supported the growth rings were formed annually, and the cross-validation study supported the quartile method (age was determined as the number of opaque bands on otolith plus the quartile of the width of the marginal translucent band) provided more appropriate estimates of age. The multimodel inference approach suggested the von Bertalanffy growth model as the optimal model for P. macrocephalus and logistic growth model for A. nibe, with asymptotic lengths and relative growth rates of 18.0 cm TL and 0.789 year−1 and 55.21 cm, 0.374 year−1, respectively. Auximetric plots of global estimates showed a downward trend with clusters by species. Growth rates of the two species were higher than in previous studies using the same aging structure (otolith) and from similar locations conducted a decade ago, suggesting a possible effect of increased fishing pressure and the need to establish a management framework. This study adds updated information to the global literature and provides an overview of growth parameters for the two important croakers.

Integrating basic and applied research to estimate carnivore abundance.

A clear connection between basic research and applied management is often missing or difficult to discern. We present a case study of integration of basic research with applied management for estimating abundance of gray wolves (Canis lupus) in Montana, USA. Estimating wolf abundance is a key component of wolf management but is costly and time intensive as wolf populations continue to grow. We developed a multimodel approach using an occupancy model, mechanistic territory model, and empirical group size model to improve abundance estimates while reducing monitoring effort. Whereas field-based wolf counts generally rely on costly, difficult-to-collect monitoring data, especially for larger areas or population sizes, our approach efficiently uses readily available wolf observation data and introduces models focused on biological mechanisms underlying territorial and social behavior. In a three-part process, the occupancy model first estimates the extent of wolf distribution in Montana, based on environmental covariates and wolf observations. The spatially explicit mechanistic territory model predicts territory sizes using simple behavioral rules and data on prey resources, terrain ruggedness, and human density. Together, these models predict the number of packs. An empirical pack size model based on 14 years of data demonstrates that pack sizes are positively related to local densities of packs, and negatively related to terrain ruggedness, local mortalities, and intensity of harvest management. Total abundance estimates for given areas are derived by combining estimated numbers of packs and pack sizes. We estimated the Montana wolf population to be smallest in the first year of our study, with 91 packs and 654 wolves in 2007, followed by a population peak in 2011 with 1252 wolves. The population declined ~6% thereafter, coincident with implementation of legal harvest in Montana. Recent numbers have largely stabilized at an average of 191 packs and 1141 wolves from 2016 to 2020. This new approach accounts for biologically based, spatially explicit predictions of behavior to provide more accurate estimates of carnivore abundance at finer spatial scales. By integrating basic and applied research, our approach can therefore better inform decision-making and meet management needs.

Input convex neural networks in nonlinear predictive control: A multi-model approach

The presented input convex neural multi-modelling approach to Model Predictive Control (MPC) has two essential advantages. Firstly, the MPC algorithm solves only convex optimisation tasks; complex not convex and multi-modal problems are not considered. Secondly, the model is trained in a simple one-step-ahead configuration and it is not used recurrently for prediction. Two neural multi-model structures are considered: fully input convex and partially input convex ones. The fully convex model structure turns out to be too restrictive for the considered polymerisation reactor benchmark process. The partially convex model leads to very good prediction accuracy and excellent control quality. Development of fully and partially input convex multi-models is thoroughly described, particularly selection of the number of the hidden nodes. Additionally, low effectiveness of a linear multi-model is discussed for comparison.

Multi-model approach to integrate climate change impact on carbon sequestration potential of afforestation scenarios in Quebec, Canada

Afforestation of unproductive or currently non-forested territories can increase carbon land sinks and thus contribute to mitigate climate change. However, investments on large-scale afforestation could be risky because of the predicted effect of climate change on forest productivity of newly created plantations. The aim of this study was to assess the carbon sequestration and mitigation potential of afforestation scenarios with different species (Picea mariana, Picea glauca, Pinus banksiana, Pinus resinosa and Populus spp) on open woodlands and abandoned farmlands in the Province of Quebec (Canada) under different radiative forcing projections. We modelled carbon dynamics in these lands under three Representative Concentration Pathways projections (RCP 2.6, RCP 4.5, and RCP 8.5) over the 2021–2100 period. The forest gap model PICUS was used to model tree growth of afforested species as a function of the Representative Concentration Pathways 2.6, 4.5 and 8.5; these data were then used as input in the Carbon Budget Model – Canadian Forest Sector 3 to simulate the evolution of ecosystem carbon stocks and fluxes as a function of forest management and climate. Carbon transfer to harvested wood products, and carbon fluxes associated with product life cycles and substitution effects on markets, were also included in the analyses. Results showed that Pinus species responded more strongly to variations in radiative forcing than for the other simulated species. Overall, aboveground biomass was particularly altered by increased radiative forcing, which in turn reduced harvesting yield and transfers to wood processing and products. At the end of the simulation, despite the expected impacts of radiative forcing on ecosystems, afforestation scenarios on open woodlands with black spruce, white spruce, and jack pine can deliver carbon mitigation of 32% – 70% over the baseline scenario and 4% – 12% for red pine on abandoned farmlands and, hence, contribute to efforts to reduce GHG emissions, especially over the long term. Although climate change is expected to impact the growth of newly planted areas as part of afforestation efforts, the results of our study suggest that the choice of species to plant and the selected forest management strategy have a greater impact on carbon stocks than climate change itself. This study provides a better understanding of the dynamics of afforestation under climate change and whether investments in plantation can contribute to GHG reduction targets.

Robust multimodel control for uncertain nonlinear MIMO systems based on ARX-Laguerre multimodel and LSDP approach

In this paper, we propose a new alternative for the robust control of nonlinear MIMO (multiple-input multiple-output) systems based on the multimodel approach and LSDP (Loop-Shaping Design Procedure)approach in the discrete case. First, the multimodel approach is exploited by representing each linear sub-model in the form of the recent linear MIMO ARX (Auto-Regressive with eXogenous inputs) -Laguerre model which guarantees a significant parameters number reduction compared with the MIMO ARX model. Thus, the resulting multivariable multimodel, entitled MIMO ARX-Laguerre multimodel, ensures a reduction of the parametric complexity with a simple recursive vector representation. However, this reduction is conditioned by an optimal choice of the pole characterising each Laguerre base. To deal with this, we propose the identification, using the genetic algorithm, of the Laguerre poles as well as the weighting function parameters of the MIMO ARX-Laguerre multimodel. This latter allows to generate a set of linear MIMO ARX-Laguerre sub-models which are exploited to propose a robust multimodel control for uncertain nonlinear MIMO systems. In fact, for each identified sub-model, we combine the LSDP approach and the RGA (Relative Gain Array) theory to develop a local robust controller. Then, a multivariable roust multimodel control algorithm is developed where the overall control strategy by multimodel approach amounts to incorporating the same weighting functions used in the considered multivariable multimodel in order to interpolate the control laws calculated from each local robust controller. The optimisation of Laguerre poles and weighting functions parameters as well as the proposed multivariable robust multimodel control algorithm are validated on the experimental prototype Quanser Aero system.

Evaluating the association between food insecurity and risk of nephrolithiasis: an analysis of the National Health and Nutrition Examination Survey.

This study aimed to investigate the relationship between self-reported food security and kidney stone formation. Data were collected from the National Health and Nutrition Examination Survey (NHANES), a database representative of the United States population. Food security status was assessed using the US Household Food Security Survey Module: Six-Item Short Form. Characteristics of patients were compared using the Chi-square test and the student t-test. Multivariate logistic regression was performed using a multi-model approach. We analyzed 6,800 NHANES survey respondents. 37.2% of respondents were categorized as having "low food security" (scores 2-4) and 24.0% having "very low food security" (scores 5-6). 8.4% of respondents had a history of kidney stones. We found that people with very low food security had a 42% increased likelihood of developing kidney stones compared to those with high or marginal food security, after controlling for race, age, and comorbidities (OR 1.42; 95% CI 1.01-1.99). Between the different food security groups, no significant differences were observed in age, race/ethnicity, body mass index, gout history, osteoporosis history, or coronary artery disease history. Lower food security was associated with slightly younger age (< 1year difference, p = 0.001), higher poverty-income ratio (p = 0.001), and many comorbidities, including kidney stones (p = 0.007). Our study provides evidence for an association between food access and the risk of kidney stone disease. Given these findings, food insecurity should be investigated as a modifiable risk factor for the development of kidney stone disease.

Environmental variability and fishing effects on artisanal flatfish fisheries along the Portuguese coast

Assessing the effects of fishing activity and environmental variability on fishing species might help to understand their fluctuations and to manage them accordingly. However, despite the increasing number of studies linking fisheries and environmental data, few had a multi-species approach considering flatfish species and different age groups. Therefore, this study aims to determine the effects of environmental and fishing related variables on commercial flatfish species across different regions of the Portuguese coast. For that, sea surface temperature, North Atlantic Oscillation, upwelling index, wind components, river discharges and fishing effort were related to landings per unit effort (LPUE) of five commercial flatfish species (Pleuronectes platessa,Dicologlossa cuneata,Microchirusspp.,Solea lascaris, andSolea solea) fished by the artisanal fleet along different regions of Portuguese coast (Northwestern coast, Southwestern coast, and Southern coast), using a multi-model statistical approach from 1989 to 2009. The fishing effort had the same effect along the life cycle and was negatively related with LPUE trends for most of the species. Regarding to environmental variability, during early phases (larval to juvenile – age-group-0) the factors that affected the recruitment, in addition to being related to the physiological needs (e.g., sea surface temperature) of individuals that will trigger growth and survival, were related to larval transport and dispersion from offshore spawning areas to coastal nursery habitats beyond (e.g., wind components and upwelling). For juvenile and young-adult phases (age-groups I and II), the effects of environmental variability appeared to be linked to physiological needs important for growth and maturation (e.g., temperature). Concluding, different effects were observed depending on the region, species, and life-cycle phase. Therefore, these environmental-fishing relationships should be used when managing local fishing resources.

Soil, Topography and Forest Structure Shape the Abundance, Richness and Composition of Fern Species in the Fragmented Tropical Landscape of Xishuangbanna, Yunnan, China

Understanding how forest fragment size, topography, forest structure, and soil properties affect plant diversity remains a crucial question in conservation biology, with ferns often being understudied. To address this knowledge gap, we surveyed the abundance, species richness, and composition of ferns in a tropical landscape in south China using 75 sites in 42 forest fragments. We then used a multi-model inference approach to assess whether fern abundance, richness, and composition were better explained by (a) fragment size, (b) topography (slope, aspect), (c) forest structure (tree basal area, light availability), or (d) soil properties (pH, Carbon, Nitrogen, Phosphorous, Calcium, Magnesium, water availability, and proportion of clay, silt, and sand). We also conducted a nestedness analysis to examine whether the composition of the fern communities in smaller fragments (0.4–1 km²) differed or represented a subset of the communities found in larger fragments (e.g., &gt;10 km²). We found that (a) fern abundance was mostly influenced by soil properties, slope, and aspect, (b) fern species richness by soil properties and slope, and (c) fern species composition by forest structure, specifically, tree basal area. We also found that fern species composition was not nested in the landscape, suggesting that smaller forest fragments had different communities from larger fragments. Our results suggest also that soil properties play an important role in maintaining fern abundance and diversity and therefore protecting soil can help conserve ferns in fragmented landscapes.

Data-driven prediction of building energy consumption using an adaptive multi-model fusion approach

This paper develops an adaptive multi-model fusion approach to predict building energy consumption, aiming to give useful suggestions for better energy control. The building energy benchmarking dataset of Chicago in 2017 is selected as the case study, where 9 features are selected as the input variables aiming to estimate the weather normalized site energy use intensity of buildings. The training dataset is clustered using the K-means algorithm and sub-models are trained based on the clustered data using the XGBoost algorithm. The sub-models are then fused by assigning a weight considering both the model reliability and the matching degree and adopting a screening algorithm to weed out the unmatching sub-models, where the influence of the threshold in the screening algorithm is studied. The root mean square error of the estimation results from a fused model is found to be 13.42 which achieves a 7.6% amelioration compared with a single model. Moreover, the adaptive multi-model fusion approach is also proved to outperform both the two-stage clustering-based regression method and the linear fusion method. Benefiting from proper treatment of samples in the fuzzy zones between clusters and the screening algorithm in the fusion process, the method proposed in our paper eventually serves as more advanced guidance in the analysis and control of building energy performance.

Assessing observed and projected flood vulnerability under climate change using multi-modeling statistical approaches in the Ouémé River Basin, Benin (West Africa)

Assessing observed and projected flood vulnerability under climate change using multi-modeling statistical approaches in the Ouémé River Basin, Benin (West Africa)

Hierarchical Attention Network of Multi-Modal Biometric for a Secure Cloud-Based User Verification

The rise of the Internet has created a need to secure verification for accessing the certain information. The cloud based biometric system can provide a convenient way to realize it. In recent years, various Artificial Intelligence models have been researched for single b iometric but rare for multi-modal biometric. This article presents new a multi-model approach as Hierarchical Attention Network (HAN) which is based on four user features i.e palm image, facial image, audio signature, and digital signature. Through the layer-by-layer filtering and improved attention mechanism, HAN can provide a higher verification accuracy, which can provide a more robust and secure user verification service. At the same time, this work also implemented the proposed model on a Cloud-based service with android application. After the evaluation, the HAN method we proposed has been proved to be better than the previous models.

Estimation and fault diagnosis for non‐linear system with time‐varying faults and measurement noises: Application on two CSTRs in series

AbstractA continuously stirred tank reactor (CSTR) is largely used in water treatment and in chemical and biological processes. It is characterized by a complex non‐linear behaviour. Operating large reactors in industry can be expensive, so a common trick used to reduce costs is to operate multiple CSTRs in series. Consequently, the CSTR is usually exposed to faults and noises. This paper addresses the design of a robust observer for estimation and fault diagnosis strategy on two CSTRs in series. The considered system is affected simultaneously by time‐varying actuator and sensor faults with measurement noises. The Takagi‐Sugeno multimodel approach is proposed to transform the non‐linear model into an interpolation of several linear sub‐models with non‐measurable premise variables. The purpose of this brief is to provide the state and the fault estimation for the considered system using a proportional multiple integral (PMI) unknown input observer. The exponential stability conditions are studied with the Lyapunov theory and L2 optimization and formulated in terms of linear matrix inequalities. In addition, a comparative study with a PMI observer is conducted. Finally, the proposed observer is used for time‐varying fault detection and isolation.

Evaluation of models for simulating soybean growth and climate sensitivity in the U.S. Mississippi Delta

Crop simulation models are indispensable tools that facilitate studies involving yield impacts, adaptation and management strategies, and policy analysis. However, uncertainty among models with respect to climate change factors varies with location according to genetic, site and management differences. This study aims to evaluate the accuracy of three process-based soybean models, GLYCIM, SoySim, and DSSAT CSM-CROPGRO-Soybean, which vary in representation of soil-water-plant components, with respect to soybean growth, development, and seed yield for multiple varieties and at different locations under current and projected climates. Experimental data from the United States Mississippi Delta region including 156 site-year-cultivar-irrigation combinations were used. Sensitivity analyses with respect to climate change scenarios involving increasing air temperature and CO2 on yield were also conducted using mean model ensemble values. Statistical criteria for estimating the goodness-of-fit of the models were Wilmott's index of agreement (IA) and root mean square error (RMSE). Simulated seed yield RMSE across all validated datasets was 0.92 Mg ha−1 or lower, with GLYCIM and CROPGRO exhibiting the best values. A similar pattern was observed for IA, which ranged between 0.82 and 0.65. Simulated yield in response to climate change factors increased by 8.8% per 100 ppm CO2 on average, and declined 4.8% per °C. Although both GLYCIM and CROPGRO models simulated yield reductions between 0.4%, and 7.2%, the SoySim model predicted positive impacts (+3.7% and + 12.1%) under two climate scenarios (∆T = 1.5 °C, CO2 = 423 ppm; ∆T = 2.0 °C, CO2 = 478 ppm). Given the difference in model structure and predictions, a multi-model ensemble approach is recommended to assess global crop production under future climate conditions along with continued improvement in model heat stress sensitivity and CO2 response. However, model improvements are still required to improve accuracy with regards to genetic x management x environmental interactions.

Multi-model ensemble approaches for simulation of evapotranspiration of karst agroforestry ecosystems

Water shortages frequently occur in karst areas, and there is an urgent need to quantify water fluxes to provide information for sustainable management of water resources. Thus, a variety of models have been developed to simulate the water balance process, including actual evapotranspiration (ETc), which is a key variable for linking water and energy cycles. However, high heterogeneity of the land surface makes it hard to get precise complete set of parameters for single model, and a single model often has uncertainties in simulating evapotranspiration in karst regions. Therefore, this study integrated three well-known individual models (Penman-Monteith, PM; Priestley and Taylor, PT; and Shuttleworth-Wallace, SW) with two multi-model ensemble approaches (Bayesian model averaging, BMA; and simple model average, SA) to enhance ETc modeling in a subtropical humid karst catchment. Results show that: 1) The individual models show different strengths for different ecosystems, which could be attributed to differences in the underlying landscape surface characteristics; 2) individual models exhibited seasonal uncertainties. For example, simulated ETc (ETs) by the PM and PT model was lower than ETo (observed ETc) during November-March but higher than during April-October for forest-grass mixed and grass ecosystems; 3) Two multi-model ensemble approaches (R2 ≥ 0.85) performed better than any individual model (R2 ≤ 0.85) most likely because multi-model ensemble approaches reduce model uncertainties by weakening the bias of individual models.

The heart in congenital diaphragmatic hernia: Knowns, unknowns, and future priorities

There is growing recognition that the heart is a key contributor to the pathophysiology of congenital diaphragmatic hernia (CDH), in conjunction with developmental abnormalities of the lung and pulmonary vasculature. Investigations to date have demonstrated altered fetal cardiac morphology, notably relative hypoplasia of the fetal left heart, as well as early postnatal right and left ventricular dysfunction which appears to be independently associated with adverse outcomes. However, many more unknowns remain, not least an understanding of the genetic and cellular basis for cardiac dysplasia and dysfunction in CDH, the relationship between fetal, postnatal and long-term cardiac function, and the impact on other parts of the body especially the developing brain. Consensus on how to measure and classify cardiac function and pulmonary hypertension in CDH is also required, potentially using both non-invasive imaging and biomarkers. This may allow routine assessment of the relative contribution of cardiac dysfunction to individual patient pathophysiological phenotype and enable better, individualized therapeutic strategies incorporating targeted use of fetal therapies, cardiac pharmacotherapies, and extra-corporeal membrane oxygenation (ECMO). Collaborative, multi-model approaches are now required to explore these unknowns and fully appreciate the role of the heart in CDH.

Electronic behavior of randomly dislocated RNA and DNA nanowires: a multi-model approach

Electronic behavior of randomly dislocated RNA and DNA nanowires: a multi-model approach

A multimodel ensemble machine learning approach for CMIP6 climate model projections in an Indian River basin

AbstractMultimodel ensemble (MME) approach would help modellers to know the advantages of individual global circulation models (GCMs) and to avoid the weaknesses associated with them, and it would help the river basin modellers to make appropriate modelling decisions. The study highlights the river basin‐scale development of MME as a convenient way to reduce the parameter and structural uncertainties in the Coupled Model Intercomparison Project Phase 6 (CMIP6) GCMs simulations after identifying the best five CMIP6 GCMs based on the rating metric calculations. Furthermore, the performance of the MME was enhanced by integrating three machine learning algorithms (artificial neural network [ANN], random forest [RF], support vector machine [SVM]). Subsequently, comparative assessment depicted the improved performance in MME‐integrated ML algorithms compared to simple arithmetic mean (SAM) in simulating observed precipitation (P), maximum temperature (Tmax), and minimum temperature (Tmin) over the Damodar River basin (DRB), India. The statistical metrics indicate that the SVM and RF methods yielded better results than SAM and ANN methods, thus selected for future projections. The robustness of the MME‐RF and MME‐SVM approach has also been observed while capturing the spatial pattern as IMD‐observed with well representation of climate indices for both wet and dry seasons. Future projections with MME‐SVM and MME‐RF suggested a possible rise in mean annual P in the range of 1.4–15% and 6.8–39% with an increasing trend in temperature (Tmax, Tmin) under the SSP245 and SSP585 scenarios, respectively. Replicating the spatial pattern of the future climatic variables projections evinced a warmer and drier climate in the southwest part of the DRB for both SSP scenarios during wet and dry season and thence warned a probable drier condition on the southwest part of the DRB in future time slices.

There’s no best model! Addressing limitations of land-use scenario modelling through multi-model ensembles

Cellular automata models are popular tools for exploring future land change pathways. But simulation modelling approaches often focus too narrowly on calibration against historic reference maps, limiting the diversity of possible outcomes. We argue that, contrary to what is commonly believed, there is no ‘best model’, and that model specification and calibration accuracy depend on the objective of the research. We propose a multi-model ensemble approach, in which a wide range of models and calibration rules sets are systematically tested against multiple metrics. We apply our approach to a case study in Spain. No single model performed well for all statistics, illustrating the danger of cherry-picking statistics for best performance. In our case study, accounting for historic land changes in model design was useful for simulating compact urban development, but limited the variability of simulation outcomes. The accessibility model driver improved urban pattern replication, while suitability without accessibility was useful for simulating low-density development encroaching on natural areas. Rather than abandoning calibrations that show low agreement with reference maps based on a small number of metrics we should seek to understand what each metric is telling us and use this information to enrich the diversity of simulated outcomes.