Pub Date : 2023-08-01DOI: 10.36334/modsim.2023.wang111
C.-H. Wang
: Estimates of extremal return levels at high average recurrence intervals (ARIs) are strongly dependent on the shape parameter of the statistical model. Because of scarcity in occurrences, however, many existing extremal data span only a few decades, often resulting in large bias and uncertainty in the estimated shape parameter of the extreme hazard model. This in turn leads to unreliable predicted extreme values at high ARIs. A common approach to ameliorate this shortcoming is the ‘super-station’ (or station-year) approach which extends the length of record and should reduce the uncertainty in high ARIs. However, the problem of predicted bias remains for return levels beyond the record length of the super-station. This paper illustrates a statistical method that provides a mechanism to obtain a hazard model that produces return levels at high ARIs with reduced bias. For an ensemble of independently collected records from a number of observational sites, the method makes use of the maximum recorded value of each of the extremal data of the ensemble, as shown in Figure 1 as the starred points (green points represent synoptic and red points non-synoptic wind gusts). The logarithmically transformed probability of the maximum recorded value at a site is shown to follow the Gumbel (Type I extreme-value) distribution, therefore multiple, say m , sites provide a sample of size m transformed probabilities of extreme values, each from a distinct site. The sample can be treated as being drawn from a Gumbel distribution, irrespective of the underlying hazard-generating mechanisms or the statistical hazard models. The method is demonstrated by an analysis of the extreme wind gust data in South Australia. The results are compared to the specifications in the Australian standard AS/NZS 1170.2:2021 and indicates that the standard may have overestimated the wind gust hazard, hence the specified design wind speeds may fall on the conservative side for South Australia.
{"title":"On reducing bias at high return levels for extreme value analysis","authors":"C.-H. Wang","doi":"10.36334/modsim.2023.wang111","DOIUrl":"https://doi.org/10.36334/modsim.2023.wang111","url":null,"abstract":": Estimates of extremal return levels at high average recurrence intervals (ARIs) are strongly dependent on the shape parameter of the statistical model. Because of scarcity in occurrences, however, many existing extremal data span only a few decades, often resulting in large bias and uncertainty in the estimated shape parameter of the extreme hazard model. This in turn leads to unreliable predicted extreme values at high ARIs. A common approach to ameliorate this shortcoming is the ‘super-station’ (or station-year) approach which extends the length of record and should reduce the uncertainty in high ARIs. However, the problem of predicted bias remains for return levels beyond the record length of the super-station. This paper illustrates a statistical method that provides a mechanism to obtain a hazard model that produces return levels at high ARIs with reduced bias. For an ensemble of independently collected records from a number of observational sites, the method makes use of the maximum recorded value of each of the extremal data of the ensemble, as shown in Figure 1 as the starred points (green points represent synoptic and red points non-synoptic wind gusts). The logarithmically transformed probability of the maximum recorded value at a site is shown to follow the Gumbel (Type I extreme-value) distribution, therefore multiple, say m , sites provide a sample of size m transformed probabilities of extreme values, each from a distinct site. The sample can be treated as being drawn from a Gumbel distribution, irrespective of the underlying hazard-generating mechanisms or the statistical hazard models. The method is demonstrated by an analysis of the extreme wind gust data in South Australia. The results are compared to the specifications in the Australian standard AS/NZS 1170.2:2021 and indicates that the standard may have overestimated the wind gust hazard, hence the specified design wind speeds may fall on the conservative side for South Australia.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"75 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":"129642405","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.ugbaje
S. Ugbaje, D. Pagendam, S. Karunaratne, T. Bishop, U. Mishra, S. Gautam, M. Farrell
: It is challenging to accurately quantify changes in soil organic carbon (SOC) stocks across landscapes and over time due to the high costs associated with soil sampling to capture inherent variabilities. While process-based models can be used to quantify management-induced changes, it is increasingly important to validate and quantify uncertainties associated with model estimates. Due to monitoring, verification, and reporting requirements under carbon accounting schemes at both project and national scales, quantifying uncertainties in SOC estimates is increasingly important. Although field SOC measurements are essential for model calibration and validation, accessing large datasets with temporally repeated measurements across the landscape is limited. As a result, there is a growing interest in using earth observation (EO) datasets to integrate and constrain model inputs/outputs to reduce uncertainties. Despite the use of surrogate information, such as EO datasets, to constrain process-based models in research, there is currently no operational framework for soil carbon models. In this study, we developed an operational framework that employs EO-derived net primary productivity and leaf area index to constrain a soil carbon model. Our case study involved using the DayCent process-based model along with field measurements from 109 sites across three catchments in New South Wales, Australia. The DayCent model is equipped with C and nitrogen cycles and biogeochemistry solutions along a 20 cm soil depth. It provides more accurate system dynamics descriptions by estimating gas exchanges of CO2, N2O, and CH4 between the soil and the atmosphere compared to RothC. The framework incorporates a Bayesian hierarchical modelling approach to account for uncertainties associated with model inputs and parameter estimates during calibration. The resulting framework is scalable, making it applicable for soil carbon projects and national-scale GHG accounting. Implementation of this framework could enhance the credibility of the Australian National Greenhouse Inventory of the land sector.
{"title":"A framework for the fusion of earth observation and soil data to constrain soil organic carbon model parameters","authors":"S. Ugbaje, D. Pagendam, S. Karunaratne, T. Bishop, U. Mishra, S. Gautam, M. Farrell","doi":"10.36334/modsim.2023.ugbaje","DOIUrl":"https://doi.org/10.36334/modsim.2023.ugbaje","url":null,"abstract":": It is challenging to accurately quantify changes in soil organic carbon (SOC) stocks across landscapes and over time due to the high costs associated with soil sampling to capture inherent variabilities. While process-based models can be used to quantify management-induced changes, it is increasingly important to validate and quantify uncertainties associated with model estimates. Due to monitoring, verification, and reporting requirements under carbon accounting schemes at both project and national scales, quantifying uncertainties in SOC estimates is increasingly important. Although field SOC measurements are essential for model calibration and validation, accessing large datasets with temporally repeated measurements across the landscape is limited. As a result, there is a growing interest in using earth observation (EO) datasets to integrate and constrain model inputs/outputs to reduce uncertainties. Despite the use of surrogate information, such as EO datasets, to constrain process-based models in research, there is currently no operational framework for soil carbon models. In this study, we developed an operational framework that employs EO-derived net primary productivity and leaf area index to constrain a soil carbon model. Our case study involved using the DayCent process-based model along with field measurements from 109 sites across three catchments in New South Wales, Australia. The DayCent model is equipped with C and nitrogen cycles and biogeochemistry solutions along a 20 cm soil depth. It provides more accurate system dynamics descriptions by estimating gas exchanges of CO2, N2O, and CH4 between the soil and the atmosphere compared to RothC. The framework incorporates a Bayesian hierarchical modelling approach to account for uncertainties associated with model inputs and parameter estimates during calibration. The resulting framework is scalable, making it applicable for soil carbon projects and national-scale GHG accounting. Implementation of this framework could enhance the credibility of the Australian National Greenhouse Inventory of the land sector.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"1 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":"129908589","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.tam
{"title":"A new Python module to convert WRF regional climate projections into CORDEX-compliant datasets","authors":"","doi":"10.36334/modsim.2023.tam","DOIUrl":"https://doi.org/10.36334/modsim.2023.tam","url":null,"abstract":"","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"11 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":"130316515","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.thomas391
{"title":"How �green� is my hydrogen? A comparative study of land, water, and wastewater footprints of renewable hydrogen production","authors":"","doi":"10.36334/modsim.2023.thomas391","DOIUrl":"https://doi.org/10.36334/modsim.2023.thomas391","url":null,"abstract":"","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"6 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":"126988493","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.singh54
D. Singh, C. Bulumulla, K. Strahan, J. Gilbert, P. Gamage, L. Márquez, V. Lemiale
: Wildfires are a serious threat in many regions of the world, including Australia. The risk of these fires is expected to continue to increase due to climate change, putting more people and communities in harm’s way. One approach to reducing the risk to lives in such fires is to plan and prepare for community evacuations. Researchers have been exploring the use of self-evacuation archetypes, clustering self-reported individual behaviours in past fires, to gain insights into who evacuates, why they do so, and when. Self-evacuation archetypes encompass a range of factors, including demographic characteristics, risk perception, social networks, and prior experience. By understanding these factors, researchers can create more realistic models of decision-making during a wildfire event. In Australia, evacuations are not mandatory, and while the understanding of the decision to leave or shelter in place has advanced, much less is understood about how these decisions play out as traffic on the transport network. For instance, intermediate trips, which are trips to destinations other than the evacuation place, can constitute a significant proportion of trips following an evacuation recommendation, and can lead to different outcomes compared to those of a coordinated evacuation. Therefore, modelling the diversity of decisions and their contribution to traffic is vital to understanding local evacuation concerns and planning safe community evacuations. In this work, we present an agent-based decision-making model and scenario for the town of Castlemaine, located in the state of Victoria, Australia. Our model is based on self-evacuation archetypes, applied to a synthetic population representing the demographics of residents of the region. The model provides a framework for understanding how different individuals are likely to respond during a wildfire event, and allows exploration of the potential impact of different interventions. We believe that our approach provides a more realistic and nuanced picture of traffic during a wildfire event and can help emergency services plan more effective response strategies.
{"title":"Modelling self-evacuation archetypes to improve wildfire evacuation traffic simulations: A regional case study","authors":"D. Singh, C. Bulumulla, K. Strahan, J. Gilbert, P. Gamage, L. Márquez, V. Lemiale","doi":"10.36334/modsim.2023.singh54","DOIUrl":"https://doi.org/10.36334/modsim.2023.singh54","url":null,"abstract":": Wildfires are a serious threat in many regions of the world, including Australia. The risk of these fires is expected to continue to increase due to climate change, putting more people and communities in harm’s way. One approach to reducing the risk to lives in such fires is to plan and prepare for community evacuations. Researchers have been exploring the use of self-evacuation archetypes, clustering self-reported individual behaviours in past fires, to gain insights into who evacuates, why they do so, and when. Self-evacuation archetypes encompass a range of factors, including demographic characteristics, risk perception, social networks, and prior experience. By understanding these factors, researchers can create more realistic models of decision-making during a wildfire event. In Australia, evacuations are not mandatory, and while the understanding of the decision to leave or shelter in place has advanced, much less is understood about how these decisions play out as traffic on the transport network. For instance, intermediate trips, which are trips to destinations other than the evacuation place, can constitute a significant proportion of trips following an evacuation recommendation, and can lead to different outcomes compared to those of a coordinated evacuation. Therefore, modelling the diversity of decisions and their contribution to traffic is vital to understanding local evacuation concerns and planning safe community evacuations. In this work, we present an agent-based decision-making model and scenario for the town of Castlemaine, located in the state of Victoria, Australia. Our model is based on self-evacuation archetypes, applied to a synthetic population representing the demographics of residents of the region. The model provides a framework for understanding how different individuals are likely to respond during a wildfire event, and allows exploration of the potential impact of different interventions. We believe that our approach provides a more realistic and nuanced picture of traffic during a wildfire event and can help emergency services plan more effective response strategies.","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"12 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":"123957675","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.weligamage
H. Weligamage, K. Fowler, M. Saft, D. Ryu, T. Peterson, M. Peel
: Recent observations suggest annual runoff can undergo prolonged shifts, resulting in severe reductions in streamflow generation compared to rainfall generation during and after multiyear droughts. Moreover, these shifts are expected to continue and possibly expand under a future drying climate. Some studies suggest that vegetation may be an important factor driving hydrological shifts under multiyear droughts (Gardiya Weligamage et al., 2023; Peterson et al., 2021). However, this hypothesis is yet to be tested more rigorously over larger areas that experienced rainfall-runoff shifts. Therefore, we investigate how vegetation responded to the Millennium Drought (from 1997 – 2009) in Victoria, Australia at different spatial scales and its relationship with hydrological changes. The
最近的观测表明，年径流可能经历长时间的变化，导致在多年干旱期间和之后产生的流量与降雨量相比严重减少。此外，这些变化预计将继续，并可能在未来的干燥气候下扩大。一些研究表明，在多年干旱条件下，植被可能是驱动水文变化的重要因素(Gardiya Weligamage等，2023;Peterson et al.， 2021)。然而，这一假设还需要在更大的地区进行更严格的测试，这些地区经历了降雨径流的变化。为此，研究了1997 - 2009年澳大利亚维多利亚州不同空间尺度下植被对千年干旱的响应及其与水文变化的关系。的
{"title":"Do vegetation changes necessarily intensify hydrological shifts under multiyear droughts?","authors":"H. Weligamage, K. Fowler, M. Saft, D. Ryu, T. Peterson, M. Peel","doi":"10.36334/modsim.2023.weligamage","DOIUrl":"https://doi.org/10.36334/modsim.2023.weligamage","url":null,"abstract":": Recent observations suggest annual runoff can undergo prolonged shifts, resulting in severe reductions in streamflow generation compared to rainfall generation during and after multiyear droughts. Moreover, these shifts are expected to continue and possibly expand under a future drying climate. Some studies suggest that vegetation may be an important factor driving hydrological shifts under multiyear droughts (Gardiya Weligamage et al., 2023; Peterson et al., 2021). However, this hypothesis is yet to be tested more rigorously over larger areas that experienced rainfall-runoff shifts. Therefore, we investigate how vegetation responded to the Millennium Drought (from 1997 – 2009) in Victoria, Australia at different spatial scales and its relationship with hydrological changes. The","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"11 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":"123332910","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.khaembah
E. Khaembah, S. Thomas, R. Cichota, J. Sharp, H. Brown
: Nitrogen (N) lost from agricultural fields to surface and groundwater systems is an important environmental problem. There is growing research interest in N management strategies to improve the sustainability of farming systems. Monitoring N balance in agricultural field is technically difficult and is complicated by differences in soil types, crops, and variability between and within seasons. Simulation modelling is an alternative approach that provides a way to evaluate mitigation options across a range of management and growing conditions. To serve as a reliable basis for nutrient management, prediction accuracy of simulations models needs to be demonstrated. This study evaluated the Simple Crop Resource Uptake Model operating within the Agricultural Production Systems sIMulator framework (SCRUM-APSIM) against field data for yield, N balance components (N uptake, soil mineral N and leaching) and soil water. Evaluation data were from a wheat-broccoli-onion crop rotation subjected to two irrigation rates (recommended and excessive) and four fertiliser N rates (N0, N1, N2, N3). No fertiliser was applied in N0, while N2 represented the recommended rate for each crop. N1 and N3 represented half and twice the rate of N2, respectively. Broccoli and onion crops were evaluated across all four fertiliser rates while a flat rate of 150 kg N/ha was applied to wheat irrespective of fertiliser N treatment. SCRUM-APSIM satisfactorily simulated crop rotations and managements as indicated by performance indices:
{"title":"Use of SCRUM-APSIM to predict soil water and soil nitrogen dynamics in arable crop rotations","authors":"E. Khaembah, S. Thomas, R. Cichota, J. Sharp, H. Brown","doi":"10.36334/modsim.2023.khaembah","DOIUrl":"https://doi.org/10.36334/modsim.2023.khaembah","url":null,"abstract":": Nitrogen (N) lost from agricultural fields to surface and groundwater systems is an important environmental problem. There is growing research interest in N management strategies to improve the sustainability of farming systems. Monitoring N balance in agricultural field is technically difficult and is complicated by differences in soil types, crops, and variability between and within seasons. Simulation modelling is an alternative approach that provides a way to evaluate mitigation options across a range of management and growing conditions. To serve as a reliable basis for nutrient management, prediction accuracy of simulations models needs to be demonstrated. This study evaluated the Simple Crop Resource Uptake Model operating within the Agricultural Production Systems sIMulator framework (SCRUM-APSIM) against field data for yield, N balance components (N uptake, soil mineral N and leaching) and soil water. Evaluation data were from a wheat-broccoli-onion crop rotation subjected to two irrigation rates (recommended and excessive) and four fertiliser N rates (N0, N1, N2, N3). No fertiliser was applied in N0, while N2 represented the recommended rate for each crop. N1 and N3 represented half and twice the rate of N2, respectively. Broccoli and onion crops were evaluated across all four fertiliser rates while a flat rate of 150 kg N/ha was applied to wheat irrespective of fertiliser N treatment. SCRUM-APSIM satisfactorily simulated crop rotations and managements as indicated by performance indices:","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"190 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":"116141632","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.kommula
S. P. Kommula, B. Lohani, D. Ryu, S. Winter
: Surface rainwater harvesting (RWH) sites gather and store rainwater that otherwise would flow into the ocean. A variety of RWH structures are employed for this purpose. Identifying a site for an RWH structure is challenging, especially in inaccessible and forested areas. Poor selection of these sites leads to wastage of resources, besides the purpose remaining unfulfilled. The surface elevation data plays a critical role among various information commonly used to find suitable locations for RWH structures. Traditionally, low-resolution digital elevation models (DEMs) have been employed for this purpose. Light Detection and Ranging (LiDAR) elevation data, characterized by higher spatial resolution and accuracy even in the presence of vegetation are becoming widely available now, showing high potential for siting these structures. This study compares the performance of LiDAR and traditionally employed low-resolution and low-accuracy DEMs (Cartosat DEM in this paper, also called CartoDEM) for siting surface RWH structures (viz Gabion and Check dam). We also analyse the effect of different LiDAR DEM resolutions on the accuracy of identifying RWH structures. An airborne LiDAR-derived DEM, originally in sub-meter resolution, is aggregated to 10-m and 30-m DEMs, which are then compared with 30-m CartoDEM for RWH siting. The criteria for selecting a RWH structure is based on the work done by Roy et.al (2022). Seven thematic layers, including runoff, lithology, soil type, geomorphology, land use, land cover, stream order, and slope, are integrated into the GIS environment using Analytical Hierarchy Process (AHP), a multi-criteria decision-making technique. A pairwise comparison is made between the seven layers and the relative weights are evaluated to prepare the suitability maps for Gabion and Check dam. The generated suitability maps at different resolutions are validated using manually identified on-ground locations across the study area. It is observed that CartoDEM misses some stream pixels, where suitable sites for Gabion and Check dam may be located. In contrast, LiDAR-derived DEMs reproduce all stream pixels, thus minimizing the chance of missing a suitable site. In addition, the stream network derived from CartoDEM shows a noticeable offset (approximately 30 m) from the on-ground stream network, which is traced manually. The locations of suitable RWH sites, generated using DEMs, are compared with reference data containing 59 field locations. The LiDAR DEMs at 10-m and 30-m resolutions report an accuracy of 95% and 81%, respectively, whereas the CartoDEM has an accuracy of 39%. Besides the poor resolution and low vertical accuracy, the non-penetration capability of optical-imagery-based DEM (CartoDEM in the present paper) is also responsible for the inferior performance. The comparison highlights the shortcomings of the low-resolution DEMs and shows the potential of LiDAR DEMs for locating suitable RWH structures even in forested areas. The outc
{"title":"Relative performance evaluation of LiDAR and Cartosat DEMs for surface rainwater harvesting site identification","authors":"S. P. Kommula, B. Lohani, D. Ryu, S. Winter","doi":"10.36334/modsim.2023.kommula","DOIUrl":"https://doi.org/10.36334/modsim.2023.kommula","url":null,"abstract":": Surface rainwater harvesting (RWH) sites gather and store rainwater that otherwise would flow into the ocean. A variety of RWH structures are employed for this purpose. Identifying a site for an RWH structure is challenging, especially in inaccessible and forested areas. Poor selection of these sites leads to wastage of resources, besides the purpose remaining unfulfilled. The surface elevation data plays a critical role among various information commonly used to find suitable locations for RWH structures. Traditionally, low-resolution digital elevation models (DEMs) have been employed for this purpose. Light Detection and Ranging (LiDAR) elevation data, characterized by higher spatial resolution and accuracy even in the presence of vegetation are becoming widely available now, showing high potential for siting these structures. This study compares the performance of LiDAR and traditionally employed low-resolution and low-accuracy DEMs (Cartosat DEM in this paper, also called CartoDEM) for siting surface RWH structures (viz Gabion and Check dam). We also analyse the effect of different LiDAR DEM resolutions on the accuracy of identifying RWH structures. An airborne LiDAR-derived DEM, originally in sub-meter resolution, is aggregated to 10-m and 30-m DEMs, which are then compared with 30-m CartoDEM for RWH siting. The criteria for selecting a RWH structure is based on the work done by Roy et.al (2022). Seven thematic layers, including runoff, lithology, soil type, geomorphology, land use, land cover, stream order, and slope, are integrated into the GIS environment using Analytical Hierarchy Process (AHP), a multi-criteria decision-making technique. A pairwise comparison is made between the seven layers and the relative weights are evaluated to prepare the suitability maps for Gabion and Check dam. The generated suitability maps at different resolutions are validated using manually identified on-ground locations across the study area. It is observed that CartoDEM misses some stream pixels, where suitable sites for Gabion and Check dam may be located. In contrast, LiDAR-derived DEMs reproduce all stream pixels, thus minimizing the chance of missing a suitable site. In addition, the stream network derived from CartoDEM shows a noticeable offset (approximately 30 m) from the on-ground stream network, which is traced manually. The locations of suitable RWH sites, generated using DEMs, are compared with reference data containing 59 field locations. The LiDAR DEMs at 10-m and 30-m resolutions report an accuracy of 95% and 81%, respectively, whereas the CartoDEM has an accuracy of 39%. Besides the poor resolution and low vertical accuracy, the non-penetration capability of optical-imagery-based DEM (CartoDEM in the present paper) is also responsible for the inferior performance. The comparison highlights the shortcomings of the low-resolution DEMs and shows the potential of LiDAR DEMs for locating suitable RWH structures even in forested areas. The outc","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"42 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":"116331571","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.nguyen328
C. Nguyen, C. W. Tan, E. Daly, Valentine Pauwels
: Observing and interpreting the flood predictions from a hydrodynamic model provides the most reliable results for connectivity analysis. However, the application of physically-based models is limited due to the complexity of their calibration, computation, and validation processes, especially when applying them to large and remote catchments with scarce temporal and spatial data. Deep learning (DL), especially Convolutional Neural Networks (CNNs), is an attractive alternative to hydrodynamic modelling. DL models can use the training data from remote sensing data to produce the results with comparably high accuracy. The DL models using remote sensing data can avoid the complicated process of setting up a hydrodynamic model, which is extremely expensive and time-consuming, especially for remote catchments. We propose an approach to manipulate the CNN models to produce a daily time series of flood extents using training data from the DEA Water Observation (https://www.dea.ga.gov.au/products/dea-water-observations) and Sentinel-2 images. The northern part of the Narran River catchment, located in the Condamine-Balonne River floodplain in New South Wales, Australia, is the showcase for this method. One-dimensional (1D) CNN (using only discharge data) and two-dimensional (2D) CNN (using discharge data and either a Digital Elevation Model or a Flood Occurrence Map) are applied. In total, for both DEA Water Observation and Sentinel-2 images, there are 440 images for training and 127 images for testing, in 21 flood events from 20/12/1987 to 31/12/2020. We conduct a detailed comparison between the two CNN structures. The 1D CNN and 2D U-Net models yielded results comparable to the satellite images with Hit Rate values of 0.853 and 0.873, respectively. The 1D CNN structure is straightforward and only requires the discharge as an input, leading to shorter computational times. The 2D CNN models allow the combination of the 2D geographic data and the spatial climate data (e.g., precipitation) in training. Therefore, the 2D CNN models result in a better prediction of flood extents. Preparing training datasets from remote sensing images for the CNN models requires fewer resources than preparing inputs for a hydrodynamic model. No bathymetric data, initial and boundary conditions are required except for the gauged flow data at the
从水动力模型观测和解释洪水预报为连通性分析提供了最可靠的结果。然而，由于其校准、计算和验证过程的复杂性，特别是在将其应用于时空数据稀缺的大型和偏远流域时，基于物理的模型的应用受到限制。深度学习(DL)，特别是卷积神经网络(cnn)，是水动力学建模的一个有吸引力的替代方案。深度学习模型可以使用来自遥感数据的训练数据来产生具有较高精度的结果。利用遥感数据建立DL模型可以避免建立水动力模型的复杂过程，该过程非常昂贵和耗时，特别是对于偏远的集水区。我们提出了一种方法来操纵CNN模型，使用来自DEA Water Observation (https://www.dea.ga.gov.au/products/dea-water-observations)和Sentinel-2图像的训练数据来产生洪水范围的每日时间序列。位于澳大利亚新南威尔士州Condamine-Balonne河漫滩的Narran河集水区北部是这种方法的展示。使用一维(1D) CNN(仅使用流量数据)和二维(2D) CNN(使用流量数据和数字高程模型或洪水发生图)。总的来说，在1987年12月20日至2020年12月31日的21次洪水事件中，对于DEA Water Observation和Sentinel-2图像，有440张图像用于训练，127张图像用于测试。我们对两种CNN结构进行了详细的比较。1D CNN和2D U-Net模型的结果与卫星图像相当，命中率分别为0.853和0.873。1D CNN结构简单，只需要放电作为输入，从而缩短了计算时间。二维CNN模型允许在训练中结合二维地理数据和空间气候数据(如降水)。因此，二维CNN模型对洪水范围的预测效果较好。从遥感图像中为CNN模型准备训练数据集比为水动力模型准备输入所需的资源更少。不需要水深数据，初始和边界条件，除了测量的流量数据
{"title":"Applications of convolutional neural networks and remote sensing data to predict flood extents","authors":"C. Nguyen, C. W. Tan, E. Daly, Valentine Pauwels","doi":"10.36334/modsim.2023.nguyen328","DOIUrl":"https://doi.org/10.36334/modsim.2023.nguyen328","url":null,"abstract":": Observing and interpreting the flood predictions from a hydrodynamic model provides the most reliable results for connectivity analysis. However, the application of physically-based models is limited due to the complexity of their calibration, computation, and validation processes, especially when applying them to large and remote catchments with scarce temporal and spatial data. Deep learning (DL), especially Convolutional Neural Networks (CNNs), is an attractive alternative to hydrodynamic modelling. DL models can use the training data from remote sensing data to produce the results with comparably high accuracy. The DL models using remote sensing data can avoid the complicated process of setting up a hydrodynamic model, which is extremely expensive and time-consuming, especially for remote catchments. We propose an approach to manipulate the CNN models to produce a daily time series of flood extents using training data from the DEA Water Observation (https://www.dea.ga.gov.au/products/dea-water-observations) and Sentinel-2 images. The northern part of the Narran River catchment, located in the Condamine-Balonne River floodplain in New South Wales, Australia, is the showcase for this method. One-dimensional (1D) CNN (using only discharge data) and two-dimensional (2D) CNN (using discharge data and either a Digital Elevation Model or a Flood Occurrence Map) are applied. In total, for both DEA Water Observation and Sentinel-2 images, there are 440 images for training and 127 images for testing, in 21 flood events from 20/12/1987 to 31/12/2020. We conduct a detailed comparison between the two CNN structures. The 1D CNN and 2D U-Net models yielded results comparable to the satellite images with Hit Rate values of 0.853 and 0.873, respectively. The 1D CNN structure is straightforward and only requires the discharge as an input, leading to shorter computational times. The 2D CNN models allow the combination of the 2D geographic data and the spatial climate data (e.g., precipitation) in training. Therefore, the 2D CNN models result in a better prediction of flood extents. Preparing training datasets from remote sensing images for the CNN models requires fewer resources than preparing inputs for a hydrodynamic model. No bathymetric data, initial and boundary conditions are required except for the gauged flow data at the","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"27 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":"121488038","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.burns
G. Burns, K. Fowler, A. Horne
: Globally, the impact of climate change on snow melt has become a growing concern. Snowfall is often a critical component of streamflow as it plays a crucial role in regulating seasonality and persistence. However, in Australia, snow cover is both intermittent and limited in space, with a lack of publicly available snow data records. As a result, snowfall is often neglected within streamflow analysis. This study presents a practical method to explicitly consider interactions with snow cover when examining risk to water supply and hydrology due to the impact of climate change. Although snow coverage in Australia is relatively small, the catchments affected are important for water supply. Previous studies have been limited in scope (e.g., focused on commercial outcomes such as changes to the ski season), and publicly available studies considering implications for water supply and hydrology are rare. The method presented here aims to be readily applicable in areas with limited data on snow fall and coverage, and within the context of broader regional water resource assessments. We integrate a snow module into a monthly rainfall-runoff modelling framework using the southern Murray-Darling Basin as a case study. Challenges include that existing snow models can be complicated, requiring data inputs such as atmospheric fluxes that are not readily available for this region. To address this, we incorporate a simple temperature-dependent threshold to determine snow cover based on an existing widely used monthly-timestep snow module (Xu et al, 1996). Calibrated to a snow extent dataset based on remote sensing imagery over 2000-2014 (Thompson, 2015), we found the method can replicate snow dynamics across several climatically-distinct snow-covered areas in south-east Australia, increasing confidence in applicability under climate change. The snow module was then added to an existing monthly rainfall runoff model (WAPABA, Wang et al, 2011). This process was applied to the Snowy Mountains (Australia’s largest alpine region) to allow a “stress test” of climate change impacts on streamflow and snow cover. As expected, results suggest that both snow coverage and duration will likely significantly reduce in the future. Results suggest
在全球范围内，气候变化对融雪的影响日益受到关注。降雪通常是河流流量的关键组成部分，因为它在调节季节性和持久性方面起着至关重要的作用。然而，在澳大利亚，积雪是断断续续的，空间有限，缺乏公开的积雪数据记录。因此，在水流分析中，降雪常常被忽略。这项研究提出了一种实用的方法，在检查由于气候变化的影响对供水和水文的风险时，明确考虑与积雪的相互作用。虽然澳大利亚的积雪面积相对较小，但受影响的集水区对供水很重要。以前的研究范围有限(例如，专注于商业结果，如滑雪季节的变化)，而且考虑到供水和水文影响的公开研究很少。本文提出的方法旨在适用于降雪量和覆盖数据有限的地区，以及更广泛的区域水资源评估。我们以墨累-达令盆地南部为例，将降雪模块整合到月降雨量-径流模型框架中。面临的挑战包括，现有的雪模式可能很复杂，需要输入大气通量等数据，而这些数据在该地区并不容易获得。为了解决这个问题，我们结合了一个简单的温度相关阈值来确定积雪覆盖，该阈值基于现有的广泛使用的月时间步积雪模块(Xu et al, 1996)。基于2000-2014年遥感影像的雪度数据集(Thompson, 2015)，我们发现该方法可以在澳大利亚东南部几个气候不同的积雪覆盖地区复制雪动态，增加了气候变化下适用性的信心。然后将雪模块添加到现有的月降雨径流模型中(WAPABA, Wang et al, 2011)。这一过程被应用于雪山(澳大利亚最大的高山地区)，以便对气候变化对河流和积雪的影响进行“压力测试”。正如预期的那样，结果表明，未来积雪面积和持续时间都可能显著减少。结果显示
{"title":"A practical approach to assessing climate change impacts on snow cover and streamflow in southeast Australia","authors":"G. Burns, K. Fowler, A. Horne","doi":"10.36334/modsim.2023.burns","DOIUrl":"https://doi.org/10.36334/modsim.2023.burns","url":null,"abstract":": Globally, the impact of climate change on snow melt has become a growing concern. Snowfall is often a critical component of streamflow as it plays a crucial role in regulating seasonality and persistence. However, in Australia, snow cover is both intermittent and limited in space, with a lack of publicly available snow data records. As a result, snowfall is often neglected within streamflow analysis. This study presents a practical method to explicitly consider interactions with snow cover when examining risk to water supply and hydrology due to the impact of climate change. Although snow coverage in Australia is relatively small, the catchments affected are important for water supply. Previous studies have been limited in scope (e.g., focused on commercial outcomes such as changes to the ski season), and publicly available studies considering implications for water supply and hydrology are rare. The method presented here aims to be readily applicable in areas with limited data on snow fall and coverage, and within the context of broader regional water resource assessments. We integrate a snow module into a monthly rainfall-runoff modelling framework using the southern Murray-Darling Basin as a case study. Challenges include that existing snow models can be complicated, requiring data inputs such as atmospheric fluxes that are not readily available for this region. To address this, we incorporate a simple temperature-dependent threshold to determine snow cover based on an existing widely used monthly-timestep snow module (Xu et al, 1996). Calibrated to a snow extent dataset based on remote sensing imagery over 2000-2014 (Thompson, 2015), we found the method can replicate snow dynamics across several climatically-distinct snow-covered areas in south-east Australia, increasing confidence in applicability under climate change. The snow module was then added to an existing monthly rainfall runoff model (WAPABA, Wang et al, 2011). This process was applied to the Snowy Mountains (Australia’s largest alpine region) to allow a “stress test” of climate change impacts on streamflow and snow cover. As expected, results suggest that both snow coverage and duration will likely significantly reduce in the future. Results suggest","PeriodicalId":390064,"journal":{"name":"MODSIM2023, 25th International Congress on Modelling and Simulation.","volume":"26 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":"124438571","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：