Development of an incorporative PSR-Fuzzy model for health assessment of the KoozehTopraghi Watershed

Introduction
Watershed degradation had negative effects on ecological and anthropologic functions at different scales. Therefore, strategic planning and conserving watershed resources is the main goal for managers and policy-makers. To achieve this goal, it is essential to provide a scientific roadmap concerning the health degree of the watershed in terms of its multi-functions. A healthy watershed improves the resilience of local ecology to climate change and provides essential services for human and ecological functions. Identifying healthy watersheds could be an effective managerial tool for monitoring natural and human phenomena and impacts. Although, in recent decades, there have been numerous types of research on watershed health and its assessment methods in different water and soil environments and in relation to environmental and social processes with economic models for decision-making in different fields. But regarding to the interpretation of different watershed health assessment models with fuzzy logic, limited studies have been carried out. This is the fact that fuzzy science has been well-considered in various sciences. In recent years, fuzzy logic has been mentioned as a powerful technique in hydrological component analysis and resource decision-making. Hydrological problems are associated with uncertainty, which is managed by fuzzy logic-based models. Fuzzy logic is based on the language of nature.  To this end, the present study was planned to accomplish our previous information on the KoozehTopraghi Watershed health and develop a new PSR-Fuzzy-based framework.
 
Materials and Methods
To do this research, firstly the pressure, state-response (PSR) model was conceptualized and customized for the study watershed. Secondly, the main criteria of road density, watershed slope, runoff coefficient, agriculture area with a slope of more than 25%, precipitation, and temperature were computed for building the pressure indicator. Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), and Enhanced Vegetation Index (EVI) also were computed for building the state indicator. Then, the specific erosion (m³ y^-1), erosion intensity coefficient, river density, and rangeland area were computed for building the response indicator. Thirdly, these criteria are converted to fuzzy bases using Fuzzy Linear membership functions in the ArcGIS 10.8 environment. Fuzzification is a method in which each pixel in the map is given a value between zero and one. This amount expresses its value according to the goal it pursues, and the higher it is in terms of value, the higher it is awarded to it as a result. Six operators including AND, OR, SUM, PRODUCT, Gamma 0.9, and Gamma 0.5 were used for incorporating three indicators of PSR and watershed health zoning. Fourthly, to evaluate and classify the output results of the operators used in the estimation of watershed health, the Quality Sum (QS) was used.
 
Results and Discussion
The results proved the better performance of two operators of Gamma 0.9 and PRODUCT. The Qs was 0.46 for PRODUCT as the first priority, followed by Gamma 0.9 operators with a Qs of 0.37 in the second priority as the most efficient operators in mapping watershed health. The pressure indicator results showed that 33.84, 0.16, 9.45, 50.51, and 6.04% of the total area of the KoozehTopraghi watershed were classified as very high, high, medium, low, and very low, respectively. The results of the state indicator, 7.55, 52.71, 39.67, and 0.07% of the total area of the study watershed were classified as very high, high, medium, and very low, respectively. The response indicator results indicated that 15.16, 13.30, 29.99, 34.80, and 6.76% of the total area of the KoozehTopraghi watershed were classified as very high, high, medium, low, and very low, respectively. According to the results of the PRODUCT operator, 67, 23, 9, and 1 % of the study watershed were classified as unhealthy, relatively unhealthy, medium, and relatively healthy, respectively. For Gamma 0.9 operator 0.9, 46, 1, 17, and 36% of watersheds were classified in unhealthy, medium, relatively healthy, and healthy classes. Based on this, it is a priority to provide suitable solutions for basic land management. Because it may be intensified the continuation of the irreparable process at the watershed level.
 
Conclusion
The results confirmed the spatial changes in health status throughout the KoozehTopraghi Watershed. Therefore, different scientific and rational programs need to be adapted to improve health to various degrees. It is highly suggested to prioritize nature-based solutions, integrated participatory management, and adaptive co-management for improving the KoozehTopraghi watershed health. Acquaintance with modern management patterns in the world, of course, with the different climatic and social conditions of our country, we can open up in the field of comprehensive watershed management compared to the past.  The watershed health index as a practical tool in watershed management can be used to determine priorities and monitor watershed status changes. In addition, since the factors affecting the management of ecosystems are considered in the health index, it can be considered as a tool for analyzing the vegetation, water, and soil resources for use with the needs of the living organism.