نشریه پژوهش های آبخیزداری
پیاپی 134 (بهار 1401)

A lack of sufficient water resources, climate change, and especially global warming, is causing apprehension in societies, particularly in watershed managers. Scientists and engineers are forecasting climatic data to prevent the waste of water resources and to optimize their use in watersheds. The Aran Station in the Kangavar Watershed was chosen to predict the trend of temperature, precipitation, and runoff using the CMIP5 climate model. Benefiting from 33 years of data (1983 to 2015), the Gene Expression Programming method (GEPM) and Decision Tree methods were developed to train, test, and predict the river discharge rate. Different climate models were implemented using the historical data of the study area. The Fgoals-g2 was chosen to predict temperature and precipitation data for the 2020-2052 periods. The RCP2.5 climate scenario was used as an optimistic scenario, and the output of the change factor downscaling method was used as an input for the GEP model and the ORANGE Software to find the best prediction of the discharge parameter in the future. The results indicated that the temperature of the next cycle will increase by 13 degrees and the maximum monthly temperature will reach 31.18 degrees centigrade. The maximum monthly precipitation will increase by 4 percent and reach 169.51mm. The longtime yearly mean precipitation will change from 423.39 mm to 427.15 mm. The correlation coefficient of the test data in the GEPM was 0.70. The maximum monthly discharge will decrease 1.84 percent, from 29.31 to 28.77 cubic meters per second (m3s-1). The mean discharge will decrease 5.71 percent, from 3.33 to 3.14 cubic meters per second. The correlation coefficient of the test data in the decision tree regression method, using the ORANGE software was 0.995. The mean discharge will increase by 10 percent and reach 3.69 m3s-1. The maximum yearly discharge will decrease by 6 percent, from 7.62 to 7.12 m3s-1.

Micro-organisms of biological soil crusts have a major role in carbon (C) cycling and atmospheric CO2 removal. Currently, the ability of erodible soils to store atmospheric C is reduced. Thus, we assessed how inoculating soil cyanobacteria onto degraded soil affects C sequestration from the atmosphere. The existing cyanobacteria (Nostoc sp., Oscillatoria sp. and Lyngbya sp.) for the C sequestration were proliferated. They were inoculated on the 40 m2 plots located in degraded land. After 83 (middle of the period) and 172 (end of the period) days, we measured the effect of this inoculation on soil organic C, calculated potential C sequestration, and potential CO2 removal from the atmosphere. The results showed that in the control treatment, the sequestrated carbon in the middle and at the end of the experiment was 0.25 and 0.27 g.m-2.day-1, respectively. Inoculation of cyanobacteria led to a significant increase of 312% and 226% of C sequestration in the middle and end of the experiment, respectively. After 172 days, the rate of C sequestration and removal of CO2 by inoculation of cyanobacteria was estimated to be 3.59 and 13.17 ton.ha-1.year-1, respectively, which was 1.10 and 4.03 ton.ha-1.year-1 in control. The results showed that inoculation of cyanobacteria by hydro-seeding technique before the wet season (early autumn) and at least 1.5 g.m-2 of biomass has the desired effect, and is also approved economically. Eventually, field-scale inoculation of cyanobacteria could be considered as a bio-based tool for sustainable development goals through carbon sequestration and rehabilitation of lands.

The process of assessing the ecological potential of the land, or measuring the potential of the land, is an important step before implementation. The Firoozabad Township is located in the Province of Fars. Irrigated farming provides the major income for this township. The main purpose of this research was to evaluate the ecological potential for irrigated agriculture in the Firoozabad Township based on the corrective approach of the current models and to present a new EMOLUP model (Eco- Socioeconomic Model of Land Use Planning) using the GIS. To evaluate the popular models, namely: the including Makhdoom, maximum limitation, arithmetic mean, EMOLUP model using the geometric mean, simple weight arithmetic mean (based on indicator and criteria), weight arithmetic mean (based on indicator and criteria) with considering a limiting factor was used for an assessment. To increase the accuracy of irrigated agriculture utilization, the indexes of water resources degradation and the drought risk were also added. The results indicated that the EMOLUP model had the highest overall accuracy (95.82), kappa coefficient (0.915), in-class coefficient (10.20), and average accuracy (96.36) among all of the used models. The simple indicator arithmetic mean and the simple weighted indicator arithmetic mean models showed the lowest overall accuracy among all methods. By adding these two indicators, using the EMOLUP model, and evaluating accuracy indicators with ground reality, it may be concluded that the process of classifying ecological potential assessment with the above method is more logical than other common methods used in this country.

Nowadays, landslides are among the geological hazards that data mining methods based on machine learning have been developed to model and predict. This paper addresses the development of a landslide susceptibility assessment that uses machine learning techniques and GIS. Artificial Neural Network (ANN) and Random Forest (RF) were compared for the landslide spatial modeling. The landslide susceptibility zoning maps consider 15 layers including geologic, morphologic, hydrologic, man-made parameters (independent variables) for landslide susceptibility assessment, and Doab Samsami watershed in Chaharmahal Bakhtiari province was chosen for the application of models due to data availability and the 174 total landslide occurrences (dependent variables). The relationship between effective factors and landslide occurrences was quantified and weighted using frequency ratio. Data independence was tested using linear multivariate regression analysis, tolerance, and VIF indices. In order to implement and validate the model, the landslide locations were randomly divided into two subsets, namely, training (70% of the total) and testing (30%), respectively. Subsequently, RF and ANN models were developed and the landslide susceptibility zonation map was produced. Maps were evaluated and validated using frequency ratio & seed cell area index, success rate, area under of receiver Operating characteristic (AUC-ROC). Results illustrated that the two factors of slope length and topographic wetness index have multicollinearity or information overlap and were removed from the modeling process in later stages. Effective factor analysis in both models showed that lithology, land use, and aspect slope factors had a significant effect on landslides, respectively. Also, the results of classification and validation of models showed that the random forest (RF) model (AUC-ROC = 0.919) was more accurate and efficient than the artificial neural network (AUC-ROC = 0.845) for landslide occurrence prediction. The results of this study can be used by executive administrations for management and planning in development and executive plans, including rural-urban development, accurate estimation in erosion models in watersheds, soil conservation, and natural resources protection.

Water has always been one of the most important pillars of life in natural and human societies. The expansion of human societies in natural systems and the loss of balance between water production and supply by watersheds to meet theirs water needs have turned the language of communication in the water sector into a language of conflict. Although human communities are increasingly trying to exploit available water resources, the importance of water and the growing need for it needs attention to be paid to the joint use and respect for each stakeholder's rights, mutual understanding, and respect for the rights of others. On the other hand, extensive studies indicate the ability of local communities and organizations to make sustainable use of shared resources and show how local groups are able to prevent the destruction of these resources by relying on their capabilities. Therefore, in this study, in order to investigate the ability of local communities to change the water governance regime downstream of the Kor River Basin, the social network analysis method and its application in determining the water governance regime was used. Macro-level indicators of the social networks, including density index and network centralization index were used. The results of the analysis of the social network of the local community of Moezabad-e-Gourgir village showed that there is considerable social capital in the connection of trust and participation in this village. On the other hand, the results of network analysis at the organizational level also showed a low level of cooperation and coordination and a high centralization of power (centralized governance regime). By integrating the local community network into the organizational network, the centralization decreased by 31% and changed the water governance regime to a fragmented regime. But due to the lack of proper relations between organizations and the local community, the level of cooperation is still low. Therefore, in order to achieve a polycentric governance regime, it is very important to increase the relations between organizations and local communities based on cooperative and participatory relations. This can increase the adaptive capacity of the socio-ecological system of the Kor river basin to change.

In this research, it is tried to identify the potential status of groundwater resources in different parts of the Telvar watershed using two machine learning models including support vector machine and random forest models. Initially, information about the wells in the region was received from the Regional Water Company of Kurdistan. The wells in the area were randomly divided into two groups of training (including 70% of data) and validation (including 30% of data). Elevation, slope, slope direction, lithology, pedology, surface curvature, land use, topographic moisture index and distance from the river were selected as predictor variables and their map was prepared in the GIS environment. The data of the training group along with the maps related to the predictor variables were entered into the support vector machine model and the random forest model. Based on the data of the training group, the parameters of the model were calibrated and adjusted and the potential of groundwater resources was predicted. The prediction accuracy of the models was determined using the statistical method of performance characteristic curve in two stages of training and validation. The results showed that accuracy of the random forest model (98.4%) was more than the support vector machine model (98.1%).

It is impossible to manage water resources in basins without the accurate determination of the peak flood discharge. Therefore, this study was carried out to determine the peak flood discharge for the return period of 10, 25, 50, and 100 years using RSM model in Darrehrood sub-basins with 12900 km2 area in Ardebil Province. Flow data of 16 hydrometric stations were collected in a statistical period of 15 years (2001-2005) and statistical deficiencies in the joint period were eliminated by using regression methods between hydrometric stations. Floods were calculated with different return periods. Also, physiographic characteristics of sub-basins that affect flood rate including area, slope, shape factor, height average, concentration time, and curve number were achieved using ArcGIS10.2 and WMS7.1 (watershed modeling system). The Goodness of Fit (R2) in the return periods of 10, 25, 50, and 100 years was estimated to be 96, 97, 95, and 94%, respectively. This indicates the model's high accuracy to predict the peak discharge in the sub-basins using the basin physiographic parameters. Model performance indexes of the model evaluation for return period of 10, 25, 50 and 100 years were calculated respectively including root mean square error (RMSE)  of 20.99, 7.16, 29.05, 39.55, relative percentage error (ε) of 52.7, 37.7, 45.1, 49.13, mean absolute error (MAE) of 0.52, 0.45, 0.48, 0.55, coefficient of residual mass (CRM) of 0.28, 0.16, 0.56, 0.82, and model efficiency (EF) of 0.74, 0.85, 0.71, 0.75. Cross-validation diagrams showed that RSM model was very suitable for the whole return periods. Paired t-test showed that the difference between predicted and actual values in different return periods was not significant (P< 0.01). The results of this study showed that the RSM has good accuracy for estimating floods in sub-basins of Ardebil Province.

The most important human impact on the hydrological cycle of watershed is land use change. Improper management of land use increases the volume of runoff and the frequency of floods. The main purpose of this study is to investigate the effect of land use change on discharge outflow in Karkhane Watershed using the Parameter Allocation Approach. Then, utilizing the philosophy of a modelling framework proposed by Bahremand (2016), we show how calibration of most model parameters can be avoided by allocating or presetting these parameters utilizing knowledge gained from sensitivity analyses, field observations and a priori specifications as a part of a parameter allocation procedure. Evaluation of the WetSpa model performance using the Allocation Parameter Approach to simulate the daily hydrograph based on the Nash-Sutcliffe and Kling-Gupta showed that the values of these criteria are 60.18 percent and 76 percent respectively. Based on this efficiency criterion, the model in the parameter allocation approach has a good performance and shows consistency in the validation period. The results showed that land use change affects the runoff of the Karkhane watershed so that in the pessimistic scenario 2, the peak discharge has increased by 15.74% and in the optimistic scenario 3; the peak discharge has decreased by 11.33.