Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.ghodrat136
M. Ghodrat, A. Simeoni
: Structure loss in wildland fires has substantially escalated during the last few decades, affected by expanded development in the countryside region, variation in fuel treatment strategies, and climate change. Wildland-urban interface (WUI) fires are a complex multi-physics problem, especially with wind direction and speed varying along natural environments. Comprehending the influence of wind speed on the behaviour of wildland fires and the resulting thermal effects is vital for accurately predicting the damage that structures may incur when exposed to such fires. This paper presents a numerical modeling approach to investigate the effect of wind speed variation on the thermal heat flux and temperature profiles of an array of structures in a typical WUI area. To simulate the effects of a wind-driven wildfire on a suburban area, nine cubic structures, each measuring 6 × 6 × 6 m, were arranged in a grid of three rows of three. The size and shape of these structures were modeled after those used in the full-scale Silsoe cube experiment (Richards and Hoxey 2012). The numerical modelling was performed using FireFOAM, a coupled fire-atmosphere model supported by a large eddy simulation (LES) solver in an open-source CFD tool called OpenFOAM. A set of two wind velocities was modelled to simulate fires burning with an intensity of 10 MW/m. The accuracy of the numerical results was confirmed by comparing them with the aerodynamic measurements of a full-scale building model under normal conditions, without the presence of fire. This analysis revealed the key physical factors that influenced the spread of the fire and its thermal effects on the buildings. The results show that at a constant fire intensity of 10 MW/m 2 , an increase in wind speed from 6 m/s to 12 m/s causes an increase in the surface temperature of all buildings, however, the temperature rise is higher on the first row of buildings compared to the second and the third row. A comparison of the temperature contours at wind speeds of 6 m/s and 12 m/s also revealed that both the average and local temperatures increased with higher wind speeds, reaching a maximum value. However, further increases in wind speed up to 12 m/s resulted in a decrease in the temperature downstream of the fire source due to convective cooling. Furthermore, the analysis of the surface temperature profile ahead of the fire front revealed that the presence of buildings has a significant impact on the development and formation of buoyant instabilities, which directly influence the behaviour of the advancing fire line. This integrated approach of fire-atmosphere modeling represents a crucial advancement in comprehending the dynamics and potential consequences of large wind-driven wildfires in the WUI region. Despite the limitations posed by experimental results in studying the effects of wind-driven wildfires on structures, the current research aims to contribute to the understanding of fire behaviour prediction in WUI. T
{"title":"Effect of wind speed on wildfire interaction with multiple structures in the wildland�urban interface","authors":"M. Ghodrat, A. Simeoni","doi":"10.36334/modsim.2023.ghodrat136","DOIUrl":"https://doi.org/10.36334/modsim.2023.ghodrat136","url":null,"abstract":": Structure loss in wildland fires has substantially escalated during the last few decades, affected by expanded development in the countryside region, variation in fuel treatment strategies, and climate change. Wildland-urban interface (WUI) fires are a complex multi-physics problem, especially with wind direction and speed varying along natural environments. Comprehending the influence of wind speed on the behaviour of wildland fires and the resulting thermal effects is vital for accurately predicting the damage that structures may incur when exposed to such fires. This paper presents a numerical modeling approach to investigate the effect of wind speed variation on the thermal heat flux and temperature profiles of an array of structures in a typical WUI area. To simulate the effects of a wind-driven wildfire on a suburban area, nine cubic structures, each measuring 6 × 6 × 6 m, were arranged in a grid of three rows of three. The size and shape of these structures were modeled after those used in the full-scale Silsoe cube experiment (Richards and Hoxey 2012). The numerical modelling was performed using FireFOAM, a coupled fire-atmosphere model supported by a large eddy simulation (LES) solver in an open-source CFD tool called OpenFOAM. A set of two wind velocities was modelled to simulate fires burning with an intensity of 10 MW/m. The accuracy of the numerical results was confirmed by comparing them with the aerodynamic measurements of a full-scale building model under normal conditions, without the presence of fire. This analysis revealed the key physical factors that influenced the spread of the fire and its thermal effects on the buildings. The results show that at a constant fire intensity of 10 MW/m 2 , an increase in wind speed from 6 m/s to 12 m/s causes an increase in the surface temperature of all buildings, however, the temperature rise is higher on the first row of buildings compared to the second and the third row. A comparison of the temperature contours at wind speeds of 6 m/s and 12 m/s also revealed that both the average and local temperatures increased with higher wind speeds, reaching a maximum value. However, further increases in wind speed up to 12 m/s resulted in a decrease in the temperature downstream of the fire source due to convective cooling. Furthermore, the analysis of the surface temperature profile ahead of the fire front revealed that the presence of buildings has a significant impact on the development and formation of buoyant instabilities, which directly influence the behaviour of the advancing fire line. This integrated approach of fire-atmosphere modeling represents a crucial advancement in comprehending the dynamics and potential consequences of large wind-driven wildfires in the WUI region. Despite the limitations posed by experimental results in studying the effects of wind-driven wildfires on structures, the current research aims to contribute to the understanding of fire behaviour prediction in WUI. T","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124099251","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.ma219
{"title":"Is global hydrological cycle accelerating at the centennial scale? A perspective from land evapotranspiration","authors":"","doi":"10.36334/modsim.2023.ma219","DOIUrl":"https://doi.org/10.36334/modsim.2023.ma219","url":null,"abstract":"","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"113 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127702288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.roberts
M. E. Roberts, Kevin Roots
: Gullies are hot spots of erosion. Gullies are the majority source of sediment that ultimately reaches the Great Barrier Reef despite occupying less than 1% of the catchment. Consequently, considerable investment and effort has focussed on preventing gully erosion through on-site remediation activities. Porous check dams (PCDs) are a common tool in erosion mitigation activities. PCDs are designed to slow the velocity of water through a channel, promoting the deposition of sediment, nutrients and seeds above the dam. Field observations suggest that, in some cases, PCDs can lead to increased scouring below the dam, risking a net increase in erosion relative to pre-intervention conditions. This paper uses the MERGE gully erosion model to explore whether the installation of a PCD can trigger increased scouring below the dam, and consequently a net increase in the amount of sediment delivered to receiving waters. Eight scenarios, covering four flow regimes and two boundary conditions, are explored. We simulate constant depth flows of 0.1 m and 0.5 m depth in a reference gully channel with inflow concentrations from the head of 50 kg/m 3 and 100 kg/m 3 . Varying depth flows are simulated with a sinusoidal function with amplitudes of 0.1 m and 0.5 m depth with the two different inflow concentrations. The reference gully is a small linear gully of 2 m width, 60 m long channel and 2% slope. The sediment is easily eroded, with a density of 1330 kg/m 3 , and 10 µ m particle size and with low cohesion. The PCD is installed 40 m from the start of the channel. The effect of the PCD is explored considering the growth of a depositional layer, and changes in the sediment delivery rate, that is the net sediment flux exiting the gully. This modelling investigation demonstrates that the installation of a PCD can lead to an internal step (or head/waterfall) forming below the PCD. In all simulations the PCD reduced the sediment delivery rate at early times, however in five of the eight scenarios the PCD resulted in a net increase in the sediment delivery rate by the end of the simulation. The increased sediment delivery rate is a direct consequence of accumulation behind the sediment creating a step, or internal head, at the PCD. This introduces an increase in the power available to erode, and therefore a greater rate of entrainment below the PCD. These results highlight the importance of ongoing monitoring and maintenance of PCDs to ensure they continue to operate as intended.
{"title":"Porous check dams and the MERGE gully erosion model","authors":"M. E. Roberts, Kevin Roots","doi":"10.36334/modsim.2023.roberts","DOIUrl":"https://doi.org/10.36334/modsim.2023.roberts","url":null,"abstract":": Gullies are hot spots of erosion. Gullies are the majority source of sediment that ultimately reaches the Great Barrier Reef despite occupying less than 1% of the catchment. Consequently, considerable investment and effort has focussed on preventing gully erosion through on-site remediation activities. Porous check dams (PCDs) are a common tool in erosion mitigation activities. PCDs are designed to slow the velocity of water through a channel, promoting the deposition of sediment, nutrients and seeds above the dam. Field observations suggest that, in some cases, PCDs can lead to increased scouring below the dam, risking a net increase in erosion relative to pre-intervention conditions. This paper uses the MERGE gully erosion model to explore whether the installation of a PCD can trigger increased scouring below the dam, and consequently a net increase in the amount of sediment delivered to receiving waters. Eight scenarios, covering four flow regimes and two boundary conditions, are explored. We simulate constant depth flows of 0.1 m and 0.5 m depth in a reference gully channel with inflow concentrations from the head of 50 kg/m 3 and 100 kg/m 3 . Varying depth flows are simulated with a sinusoidal function with amplitudes of 0.1 m and 0.5 m depth with the two different inflow concentrations. The reference gully is a small linear gully of 2 m width, 60 m long channel and 2% slope. The sediment is easily eroded, with a density of 1330 kg/m 3 , and 10 µ m particle size and with low cohesion. The PCD is installed 40 m from the start of the channel. The effect of the PCD is explored considering the growth of a depositional layer, and changes in the sediment delivery rate, that is the net sediment flux exiting the gully. This modelling investigation demonstrates that the installation of a PCD can lead to an internal step (or head/waterfall) forming below the PCD. In all simulations the PCD reduced the sediment delivery rate at early times, however in five of the eight scenarios the PCD resulted in a net increase in the sediment delivery rate by the end of the simulation. The increased sediment delivery rate is a direct consequence of accumulation behind the sediment creating a step, or internal head, at the PCD. This introduces an increase in the power available to erode, and therefore a greater rate of entrainment below the PCD. These results highlight the importance of ongoing monitoring and maintenance of PCDs to ensure they continue to operate as intended.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126288054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.schopf
{"title":"A guide to future climate projections for water resource management in Western Australia","authors":"","doi":"10.36334/modsim.2023.schopf","DOIUrl":"https://doi.org/10.36334/modsim.2023.schopf","url":null,"abstract":"","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126359962","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.holzworth
D. Holzworth, N. Huth
: Modelling at scale involves creating workflows that connect data to tools, utilities, and models. Often this is a manual process (e.g. scripts with no automation) that evolves over time. Unless there is clear, detailed documentation, that is accessible, it can be very difficult to reproduce simulation results at some point in the future. Journal paper descriptions of simulation results are often not reproducible! The software development industry created Docker images to very clearly define an execution environment that is reproducible. The docker user creates a simple text-based recipe (dockerfile) that installs the software application (model) and its dependencies into an image that can be executed repeatedly. If the image is pushed to a docker repository (e.g. DockerHub) then it will be accessible by others. This solves part of the reproducibility problem by encapsulating the execution environment into a sharable image. It doesn’t solve the problem of identifying the model input data.
{"title":"Reproducible modelling: Why is it so hard?","authors":"D. Holzworth, N. Huth","doi":"10.36334/modsim.2023.holzworth","DOIUrl":"https://doi.org/10.36334/modsim.2023.holzworth","url":null,"abstract":": Modelling at scale involves creating workflows that connect data to tools, utilities, and models. Often this is a manual process (e.g. scripts with no automation) that evolves over time. Unless there is clear, detailed documentation, that is accessible, it can be very difficult to reproduce simulation results at some point in the future. Journal paper descriptions of simulation results are often not reproducible! The software development industry created Docker images to very clearly define an execution environment that is reproducible. The docker user creates a simple text-based recipe (dockerfile) that installs the software application (model) and its dependencies into an image that can be executed repeatedly. If the image is pushed to a docker repository (e.g. DockerHub) then it will be accessible by others. This solves part of the reproducibility problem by encapsulating the execution environment into a sharable image. It doesn’t solve the problem of identifying the model input data.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125617736","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.nguyen546
H. Nguyen, N. Tuteja, Hemantha Perera, A. Raut, Tahir Hameed, Richa Neupane, A. Breda
: WaterNSW supplies bulk water to its customers and operates a large network of dams and rivers of NSW. For operations, we use the river system models within CARM that are based on daily and hourly scales for river operations, and flood and storage operations. These models use deterministic approaches for streamflow predictions rather than probabilistic framework. Considerable advances in probabilistic risk-based hydrologic and hydroclimate modelling have been made in research and operational settings over the last decade nationally and overseas (Bennett et al., 2014; McInerney et al., 2020). We investigate the performance of daily GR4J model with the choice of different objective functions for use in probabilistic forecasting. GR4J model is chosen for its effectiveness in real-time operational forecasting applications, owing to its simplicity, computational efficiency, and lower data requirements. It has also been tested and used in many streamflow forecasting agencies in France, Australia, and other countries. The three objective functions chosen for this investigation include SDEB (Square-root Daily, Exceedance and Bias) in Source (eWater) generally used for river system planning models, NSE-BC0.2 (Nash-Sutcliffe Efficiency with Box-Cox Transformation set to 0.2) in the Multi-Temporal Hydrological Residual Error (MuTHRE) model used for seasonal streamflow forecasts, and NSE-SCHEF in SWIFT (Short-term Water Information and Forecasting Tools) used for 7-day streamflow forecasts. We investigate how well the three objective functions perform using a deterministic performance evaluation criterion covering low-, medium-to high-flow range. Seven catchments in Lachlan, Namoi Peel and Murrumbidgee are selected for this investigation. A leave-one-year-out cross-validation approach is implemented for all the seven catchments. Some of the typical results are provided in Figure 1 for reference. The
:新南威尔士州水务公司为其客户提供大量的水，并在新南威尔士州经营着一个由水坝和河流组成的大型网络。对于操作，我们使用CARM中的河流系统模型，该模型基于河流操作、洪水和储存操作的每日和小时尺度。这些模型使用确定性方法进行流量预测，而不是概率框架。在过去十年中，基于概率风险的水文和水文气候模型在国内外的研究和操作环境中取得了相当大的进展(Bennett等人，2014;McInerney et al.， 2020)。我们通过选择不同的目标函数来研究每日GR4J模型在概率预测中的性能。选择GR4J模型是因为其简单、计算效率高、数据要求低，在实时业务预测应用中具有有效性。它还在法国、澳大利亚和其他国家的许多流量预报机构中进行了测试和使用。本研究选择的三个目标函数包括用于河流系统规划模型的源(eWater)中的SDEB(平方根日、超标和偏差)，用于季节性流量预测的Multi-Temporal水文残差(MuTHRE)模型中的NSE-BC0.2 (Box-Cox变换设为0.2的Nash-Sutcliffe效率)，以及用于7天流量预测的SWIFT(短期水信息和预测工具)中的NSE-SCHEF。我们使用涵盖低、中、高流量范围的确定性性能评估标准来研究这三个目标函数的表现。在拉克兰，纳莫伊皮尔和Murrumbidgee的七个集水区被选为这次调查的对象。对所有七个集水区实施了留出一年的交叉验证方法。图1中提供了一些典型的结果以供参考。的
{"title":"The influence of different objective functions in GR4J model-on-model performance for streamflow forecasting application","authors":"H. Nguyen, N. Tuteja, Hemantha Perera, A. Raut, Tahir Hameed, Richa Neupane, A. Breda","doi":"10.36334/modsim.2023.nguyen546","DOIUrl":"https://doi.org/10.36334/modsim.2023.nguyen546","url":null,"abstract":": WaterNSW supplies bulk water to its customers and operates a large network of dams and rivers of NSW. For operations, we use the river system models within CARM that are based on daily and hourly scales for river operations, and flood and storage operations. These models use deterministic approaches for streamflow predictions rather than probabilistic framework. Considerable advances in probabilistic risk-based hydrologic and hydroclimate modelling have been made in research and operational settings over the last decade nationally and overseas (Bennett et al., 2014; McInerney et al., 2020). We investigate the performance of daily GR4J model with the choice of different objective functions for use in probabilistic forecasting. GR4J model is chosen for its effectiveness in real-time operational forecasting applications, owing to its simplicity, computational efficiency, and lower data requirements. It has also been tested and used in many streamflow forecasting agencies in France, Australia, and other countries. The three objective functions chosen for this investigation include SDEB (Square-root Daily, Exceedance and Bias) in Source (eWater) generally used for river system planning models, NSE-BC0.2 (Nash-Sutcliffe Efficiency with Box-Cox Transformation set to 0.2) in the Multi-Temporal Hydrological Residual Error (MuTHRE) model used for seasonal streamflow forecasts, and NSE-SCHEF in SWIFT (Short-term Water Information and Forecasting Tools) used for 7-day streamflow forecasts. We investigate how well the three objective functions perform using a deterministic performance evaluation criterion covering low-, medium-to high-flow range. Seven catchments in Lachlan, Namoi Peel and Murrumbidgee are selected for this investigation. A leave-one-year-out cross-validation approach is implemented for all the seven catchments. Some of the typical results are provided in Figure 1 for reference. The","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131386152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.ren355
{"title":"A rapid analytical model to represent dual-priority water rights in carryover systems","authors":"","doi":"10.36334/modsim.2023.ren355","DOIUrl":"https://doi.org/10.36334/modsim.2023.ren355","url":null,"abstract":"","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"95 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132035128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.zvezdin
Andrei V. Zvezdin, Mitchell Welch, T. Schaerf
: Insect swarms, fish schools and bird flocks are familiar everyday examples of collective motion. How these groups synchronize despite being made up of vastly different individuals is a fascinating question. Classical modeling of collective motion assumes homogenous individuals interacting locally to produce group-level patterns which are then classified via summary measures, usually called order parameters, such as the degree of alignment or rotation of the group. Empirical and theoretical work have pointed to the importance of individual differences (heterogeneity) driving collective behaviour. To investigate how individual differences drive collective motion, we need to understand how the different members contribute to the emergent collective motion phenomena. For this we need to shift the focus from the group to the individual and either introduce new measures or generalize the group level order parameters to the individual level. We investigated the following measures for studying heterogeneous groups: 1) the individual state (cycling, directed, random or composite trajectories) derived from applying alignment and milling order parameters to an individual’s track; 2) the individual fluidity defined as the amount of movement relative to the group centroid over a fixed p eriod; 3 ) t he d ichotomy d efined as th e lo cal di fference in he ading be tween a focal individual and its neighbours and 4) the number of neighbours an individual sees or interacts with. Using a canonical zonal (repulsion, orientation, attraction) constant speed self propelled particle model we applied the above measures to homogenous and heterogeneous groups comprised of two distinct behavioural class types to investigate key questions of multi-stability, self-sorting and state transitions. We investigated these questions in the context of hysteresis as it is a natural measure of multi-stability and the tendency of a collective motion system to exhibit a form of collective memory where current emergent group behaviour is influenced by the recent history of the system. We produced hysteresis loops for heterogeneous groups by keeping one subgroup
{"title":"Individual measures for heterogeneous collective motion","authors":"Andrei V. Zvezdin, Mitchell Welch, T. Schaerf","doi":"10.36334/modsim.2023.zvezdin","DOIUrl":"https://doi.org/10.36334/modsim.2023.zvezdin","url":null,"abstract":": Insect swarms, fish schools and bird flocks are familiar everyday examples of collective motion. How these groups synchronize despite being made up of vastly different individuals is a fascinating question. Classical modeling of collective motion assumes homogenous individuals interacting locally to produce group-level patterns which are then classified via summary measures, usually called order parameters, such as the degree of alignment or rotation of the group. Empirical and theoretical work have pointed to the importance of individual differences (heterogeneity) driving collective behaviour. To investigate how individual differences drive collective motion, we need to understand how the different members contribute to the emergent collective motion phenomena. For this we need to shift the focus from the group to the individual and either introduce new measures or generalize the group level order parameters to the individual level. We investigated the following measures for studying heterogeneous groups: 1) the individual state (cycling, directed, random or composite trajectories) derived from applying alignment and milling order parameters to an individual’s track; 2) the individual fluidity defined as the amount of movement relative to the group centroid over a fixed p eriod; 3 ) t he d ichotomy d efined as th e lo cal di fference in he ading be tween a focal individual and its neighbours and 4) the number of neighbours an individual sees or interacts with. Using a canonical zonal (repulsion, orientation, attraction) constant speed self propelled particle model we applied the above measures to homogenous and heterogeneous groups comprised of two distinct behavioural class types to investigate key questions of multi-stability, self-sorting and state transitions. We investigated these questions in the context of hysteresis as it is a natural measure of multi-stability and the tendency of a collective motion system to exhibit a form of collective memory where current emergent group behaviour is influenced by the recent history of the system. We produced hysteresis loops for heterogeneous groups by keeping one subgroup","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"117 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129969287","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.khorshidi
M. S. Khorshidi, M. Gandomi, M. Nikoo, D. Yazdani, Fang Chen, A. Gandomi
: Genetic programming (GP) has shown great promise in empirical modelling for environmental science, particularly in complex systems such as climate, flood, and environmental modelling. However, the success of GP largely depends on the quality and quantity of data used for training. In this regard, knowledge discovery (KD) can significantly improve GP’s ability to model complex interactions (Grin and Gandomi 2021). KD is the process of discovering new knowledge or insights from existing data, often through data mining and machine learning techniques. KD can be used in conjunction with GP to identify relevant variables, patterns, and interactions within a dataset, which can then be used to improve the accuracy and generalization of GP models. By discovering new knowledge, KD can also help GP to avoid overfitting and capture more complex relationships between variables.
遗传规划(GP)在环境科学的经验建模中显示出巨大的前景，特别是在复杂系统中，如气候、洪水和环境建模。然而，GP的成功在很大程度上取决于用于训练的数据的质量和数量。在这方面，知识发现(KD)可以显著提高GP对复杂交互建模的能力(Grin and Gandomi 2021)。KD是从现有数据中发现新知识或见解的过程，通常通过数据挖掘和机器学习技术。KD可以与GP结合使用，以识别数据集中的相关变量、模式和相互作用，然后可用于提高GP模型的准确性和泛化。通过发现新知识，KD还可以帮助GP避免过拟合，并捕获变量之间更复杂的关系。
{"title":"Enhancing empirical modelling in environmental science with knowledge discovery and genetic programming","authors":"M. S. Khorshidi, M. Gandomi, M. Nikoo, D. Yazdani, Fang Chen, A. Gandomi","doi":"10.36334/modsim.2023.khorshidi","DOIUrl":"https://doi.org/10.36334/modsim.2023.khorshidi","url":null,"abstract":": Genetic programming (GP) has shown great promise in empirical modelling for environmental science, particularly in complex systems such as climate, flood, and environmental modelling. However, the success of GP largely depends on the quality and quantity of data used for training. In this regard, knowledge discovery (KD) can significantly improve GP’s ability to model complex interactions (Grin and Gandomi 2021). KD is the process of discovering new knowledge or insights from existing data, often through data mining and machine learning techniques. KD can be used in conjunction with GP to identify relevant variables, patterns, and interactions within a dataset, which can then be used to improve the accuracy and generalization of GP models. By discovering new knowledge, KD can also help GP to avoid overfitting and capture more complex relationships between variables.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134367100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.stenborg
T. Stenborg
: Stan is a probabilistic programming language that uses Markov chain Monte Carlo (MCMC) sampling for Bayesian inference (Carpenter at el.). Stan sampling can be parallelised by running Markov chains m on separate processing cores n , i.e. ≥ 1 chain/core, for Amdahlian speedup (Annis et al.). An extension , introduced here, is adaptive parallelisation. First, prior to planned sampling, performance benchmarking was dynamically performed with m = 4… M chains distributed over n = 1 … m cores (where M is a system’s number of available cores, and using at least four chains is recommended (Vehtari et el.)). The best performing configuration ( m, n ) was then automatically adopted ( github.com/tstenborg/Stan - Adaptive -Parallelisation). To be relevant, benchmarking should proceed with the same data and compiled Stan model as the planned sampling. For efficiency, benchmarking was performed with fewer chain iterations than for inference proper, though using the same ratio of warmup to post-warmup iterations/chain (1 : 1/ m , yielding an equal number of total draws per configuration). For further efficiency, comparison of only one evaluation of each configuration was made. One evaluation was deemed sufficient after measuring speedup variability, for an example problem and configuration near the middle of a test system’s (Intel Core i7-10750H) non-hyperthreaded ( m , n ) configuration range. The simplifying assumption was made that results for the configuration were representative of the entire hyperthreaded and non-hyperthreaded range. Finally, for meaningful interconfiguration comparisons, a fixed seed was passed to the Stan random number generator. Warmup iterations had a significant effect on optimum ( m , n ). Too few warmup iterations, though speeding up benchmarking, can leave Stan without enough adaptation time to determine efficient sampling parameters (Hecht et al.
{"title":"Adaptive MCMC parallelisation in Stan","authors":"T. Stenborg","doi":"10.36334/modsim.2023.stenborg","DOIUrl":"https://doi.org/10.36334/modsim.2023.stenborg","url":null,"abstract":": Stan is a probabilistic programming language that uses Markov chain Monte Carlo (MCMC) sampling for Bayesian inference (Carpenter at el.). Stan sampling can be parallelised by running Markov chains m on separate processing cores n , i.e. ≥ 1 chain/core, for Amdahlian speedup (Annis et al.). An extension , introduced here, is adaptive parallelisation. First, prior to planned sampling, performance benchmarking was dynamically performed with m = 4… M chains distributed over n = 1 … m cores (where M is a system’s number of available cores, and using at least four chains is recommended (Vehtari et el.)). The best performing configuration ( m, n ) was then automatically adopted ( github.com/tstenborg/Stan - Adaptive -Parallelisation). To be relevant, benchmarking should proceed with the same data and compiled Stan model as the planned sampling. For efficiency, benchmarking was performed with fewer chain iterations than for inference proper, though using the same ratio of warmup to post-warmup iterations/chain (1 : 1/ m , yielding an equal number of total draws per configuration). For further efficiency, comparison of only one evaluation of each configuration was made. One evaluation was deemed sufficient after measuring speedup variability, for an example problem and configuration near the middle of a test system’s (Intel Core i7-10750H) non-hyperthreaded ( m , n ) configuration range. The simplifying assumption was made that results for the configuration were representative of the entire hyperthreaded and non-hyperthreaded range. Finally, for meaningful interconfiguration comparisons, a fixed seed was passed to the Stan random number generator. Warmup iterations had a significant effect on optimum ( m , n ). Too few warmup iterations, though speeding up benchmarking, can leave Stan without enough adaptation time to determine efficient sampling parameters (Hecht et al.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128928753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：