Volume 18, Issue 2 (7-2024)
مرتع 2024, 18(2): 224-240 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Amiri Largani S, jafarian Z, Rastgar S. Ecotourism Potential Modeling Using Non-Parametric Algorithms in Western Rangelands of Mazandaran Province. مرتع 2024; 18 (2) :224-240
URL: http://rangelandsrm.ir/article-1-1182-en.html
URL: http://rangelandsrm.ir/article-1-1182-en.html
Department of Range Management, Faculty of Range and Watershed Management, Sari University of Agricultural Sciences and Natural Resources, Sari
Abstract: (674 Views)
Background and objectives: Ecotourism is a powerful approach to economic development and nature conservation. It induces positive behavioral changes in both local residents and tourists, fostering the preservation of natural and cultural resources. This study aims to model and rank areas with ecotourism potential in the rangelands of Mazandaran province.
Methodology: Utilizing Makhdoom's system method, the study evaluates the ecotourism potential of Western Rangelands in Mazandaran province. Environmental and ecological factors were identified and mapped. These maps were then analyzed, summarized, and integrated using GIS, based on the Makhdoom model, to classify the area's recreational potential. Modeling involved correlating recreational potential with environmental and recreational factors using data mining techniques: artificial neural networks (ANN), random forests, and support vector machines (SVM), facilitated by STATISTICA software. Independent variables included slope, direction, elevation, soil, vegetation, water resources, access routes, lithology, distance to faults, power lines, and proximity to urban areas. The dependent variable was ecotourism potential. Map preparation, modification, and classification were performed in ArcGIS 10.8.2, with statistical data analysis in Excel.
Results: According to the final recreation plan, Chenarbon rangeland with an area of 292/1514 hectares has the possibility of extensive recreation of the second floor. In Dasht-Lashak, 46.51% has the possibility of extensive recreation on the second floor and 53.49% is unsuitable for recreation, also in the rangeland of Islamabad, 75.2% has the potential for extensive recreation on the second floor and 24.98% is unsuitable for recreation. According to the final recreation map, 96.19% of the second floor has extensive recreation capacity in the Kohneh-Lashak rangeland and 3.81% is unsuitable for recreation. Kojour rangeland with 87.24% has the possibility of extensive recreation of the second floor and 12.76% of the rangeland is unsuitable for recreation. Largan rangeland with 12.54% has the capacity of extensive recreation of the second floor and 87.46% is unsuitable for recreation, and finally, Piedeh rangeland with 57.31% has the capacity for extensive recreation of the second floor and 42.69% is unsuitable for recreation. 65.83% of the area is suitable for recreation and 34.17% of the area is unsuitable for recreation. Based on the value of the explanation coefficient, the most appropriate model among the three different algorithms investigated is the ANN model with the BFGS algorithm with the 8-10-1 MLP network with the number of 8 input layers, 10 hidden layers, and one output layer in modeling and (R2=0.94, RMSE=9.99% and BIAS=0.02%) were selected.
Conclusion: The study confirms that non-parametric algorithms provide highly accurate evaluations of ecotourism potential. It is recommended to use data mining techniques to design ecotourist services, meeting their basic needs while preserving nature.
Methodology: Utilizing Makhdoom's system method, the study evaluates the ecotourism potential of Western Rangelands in Mazandaran province. Environmental and ecological factors were identified and mapped. These maps were then analyzed, summarized, and integrated using GIS, based on the Makhdoom model, to classify the area's recreational potential. Modeling involved correlating recreational potential with environmental and recreational factors using data mining techniques: artificial neural networks (ANN), random forests, and support vector machines (SVM), facilitated by STATISTICA software. Independent variables included slope, direction, elevation, soil, vegetation, water resources, access routes, lithology, distance to faults, power lines, and proximity to urban areas. The dependent variable was ecotourism potential. Map preparation, modification, and classification were performed in ArcGIS 10.8.2, with statistical data analysis in Excel.
Results: According to the final recreation plan, Chenarbon rangeland with an area of 292/1514 hectares has the possibility of extensive recreation of the second floor. In Dasht-Lashak, 46.51% has the possibility of extensive recreation on the second floor and 53.49% is unsuitable for recreation, also in the rangeland of Islamabad, 75.2% has the potential for extensive recreation on the second floor and 24.98% is unsuitable for recreation. According to the final recreation map, 96.19% of the second floor has extensive recreation capacity in the Kohneh-Lashak rangeland and 3.81% is unsuitable for recreation. Kojour rangeland with 87.24% has the possibility of extensive recreation of the second floor and 12.76% of the rangeland is unsuitable for recreation. Largan rangeland with 12.54% has the capacity of extensive recreation of the second floor and 87.46% is unsuitable for recreation, and finally, Piedeh rangeland with 57.31% has the capacity for extensive recreation of the second floor and 42.69% is unsuitable for recreation. 65.83% of the area is suitable for recreation and 34.17% of the area is unsuitable for recreation. Based on the value of the explanation coefficient, the most appropriate model among the three different algorithms investigated is the ANN model with the BFGS algorithm with the 8-10-1 MLP network with the number of 8 input layers, 10 hidden layers, and one output layer in modeling and (R2=0.94, RMSE=9.99% and BIAS=0.02%) were selected.
Conclusion: The study confirms that non-parametric algorithms provide highly accurate evaluations of ecotourism potential. It is recommended to use data mining techniques to design ecotourist services, meeting their basic needs while preserving nature.
Type of Study: Research |
Subject:
General
Received: 2023/01/6 | Accepted: 2023/09/26 | Published: 2024/07/31
Received: 2023/01/6 | Accepted: 2023/09/26 | Published: 2024/07/31
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
