FOREST ROADs are constructed to facilitate FOREST protection, reFORESTation, logging operations and maximizing the value of FOREST products. Therefore FOREST ROADs are key infrastructures in the development of the region. This study aims to plan FOREST ROAD NETWORK using artificial neural NETWORK and GIS regarding FOREST ROAD technical principles. First the criteria were chosen using Delphi method and then they were weighted regarding their importance in ROAD planning. After that the criteria were combined with corresponding weighs to achieve suitability map based on the degree of suitability for ROAD allocation. Value and coordinates of each pixel were extracted by ENVI software and were normalized in the range of 0-1 for modeling by MATLAB software. In this study two neural NETWORKs were used for modeling, including multilayer perceptron (MLP) and radial-bases functions (RBF). The neural NETWORKs estimated suitability of different pixels in other districts based on the Patom district results. Using an ArcView GIS extension, PEGGER, two FOREST ROAD NETWORKs were planned. The results showed that MLP provides better ability for estimating suitability of pixels for ROAD passage in comparison with RBF with the R2 of 0. 994. A linear regression was also used to compare the results of the proposed neural NETWORKs. The results revealed that neural NETWORK improves the results in comparison with the linear regression and results showed that the second ROAD alternative was optimum NETWORK with regard to the unit cost.

FOREST ROAD NETWORKs are essential structures to achieve the FORESTry aims, but these structures include most implemental and environmental costs. To minimize these costs, the effective factors have to be considered in FOREST ROAD design which landslide susceptibility is one of the major factors. ROAD construction on susceptible terrain increase ROAD maintenance cost and environmental impacts. Landslide hazard zonation is a method that can be applied to recognize and avoid the unstable terrain. In this research, the landslide hazard zonation map of Baharbon district in Kheyroud Research FOREST was developed using a combined model, AHP and Density Area methods. First, Landslides distribution map of study area was prepared in ArcGIS using field observations and then the map of effective factors in landslide occurrence were prepared. To determine the weight of effective factors AHP was used. Afterward a bivariate statistical model, Density Area, was used for calculating the weight of classes of effective factors maps. By combining the effective factors maps and classifying the combined map with regard to breakpoints of frequency curve, the landslide hazard zonation map was developed. According to the results, 14.7%, 26.9%, 38.9%, 15% and 4.5% of the district were classified as very low, low, moderate, high and very high hazard, respectively. Finally due to lack of landslide occurrence in slope class of 0-15%, as well as it is suitable slope for ROAD design, this class was suggested as positive control points. The other slope classes that were in low and very low hazard were defined as second priority for ROAD design. ROAD building in high and very high hazard is possible, if increasing the number of culverts, also biological and mechanical reinforcements of cut and fill slopes.

A study was conducted to determine the most appropriate FOREST ROAD NETWORK using GIS software at series7, district 38 of Sardabrood and Chalus in 2013. To investigate and design FOREST ROADs, a land digitalized model in TIN format was prepared by 3D map of FOREST organization in a scale of 1:25000. Then, basic maps such as maps of slope, aspect and contour line distance of 10-m were prepared using this model. In the next step, the map of geology, penology, typology, proposed paths and parcel borders were digitalized based on the booklet of FOREST management plan. Eight effective factors in designing of FOREST ROAD NETWORKs were investigated such as FOREST type, geology, pedology, distance from ROAD, hydrography NETWORK, geographical aspects of the area, slope, and landscapes. The effect of each factor was weighed using hierarchical analysis. Then in the GIS environment considering the overlapping of aforementioned eight factors, capability ROAD pass map of the area was created. Three variants of various paths in the ROAD quickly predicted and evaluated by inserting the desired longitudinal slope (regarding to the kind of ROADs) into PEGGER program. The evaluation of the path proposed by the organization and the designed paths was done using Backmond regulations and capability ROAD pass maps. The results showed that the proposed path with the density of 9.39 m per ha and coverage of 85.33% had the most suitable FOREST ROAD NETWORK.

FORESTs as a renewable source has effective and useful role in economic and social activities of society. NETWORK of FOREST ROADs as infrastructure division’s FOREST planning units (compartments) of FOREST management plans, management, logging, exit of Wood and plays a major role FOREST protection. Aim of this study, designed FOREST ROAD NETWORK according to multi-purpose FORESTry using the NETWORK analysis model in the one series of Darabkla FOREST. For this purpose, initially, the effective criteria and alternatives in the design of FOREST ROAD NETWORK in the area Determined, and then the NETWORK between criteria was designed. In the next step, using the expertise and ANP method, determined the final weight of effective alternatives in designing FOREST ROAD NETWORK and maps required for this evaluation were provided. Then using weighted linear combination (WLC) potential final map was provided for ROAD design. Then using the software Pegger was designed FOREST ROAD NETWORK. Comparison of existing ROADs and proposed ROADs done by selecting 30 points in place of the two ROADs and Chi-square test. The results showed that the five classes of cross ROADs making plans for construction ROAD has significant difference at the level of 95%. In overall, the results of this study showed that the use ANP model with acceptable performance and has made a significant difference to the existing NETWORK.

FOREST ROAD NETWORK as the basic installations has a major role in organization the region, harvesting and transportation goods and services and keeping it. The ROADs NETWORK from planning single route to completing it has technical and principle points that needs to make an exact decision. Up to now most of the designs have made based on economic and harvest goals.GIS-based planning methods have the capabilility of saving and analyzaing a great deal of data as different digital layers with high precision and speed. So it can improve planning process. One of the goals of this study is to develop a method to design a FOREST ROADs NETWORK based environmental issues and technical considerations. First, the necessary data and map and positive or negative points were extracted in the form of information layers (layer, Slop, Direction, Soil stability, Type, Volume per hectare and so on). Next so that after classifying and valuating the layers of the maps we used pair- wise comparison methods. Then weights, factors, and constraints were entered to create final suitability map and the layers were overlaperd due to important coefficient of layers based on using of Arc GIS software. In the next stage six variants of the ROAD NETWORK was designed using PEGGER software so that in the all of the variants, we tried the routs cross high levels of values. Finally, we used technical (BACKMOND, SEGBADEN, TON*Km AND Corrected TON*Km) assessment methods and their combination to identify the optimum route (the six variant).At the same time as a result it was obvious that due to a large amount of data. The designing using GIS has advantages than another methods because of precision, quality, expense and speed analyzing.

Background and Objective: FOREST ROADs are important components of FOREST management units and involving special roles in wood transporting and the other FOREST uses such as hunting, tourism, etc. One of the most important problems due to FOREST ROADs making is increasing of erosion different forms in around of ROADs, sediment yield resulting of vegetation destroy, soil structure ruin, run off velocity increasing and unstable of slope.Method: In this research, sediment due to existence FOREST ROADs was estimated in a part of Experimental FOREST Station of Tarbiat Modares University using of SEDMODEL and GIS. In this model effective factors consist of geology, ROAD, slope, precipitation, vegetation cover and drainage NETWORK.Discussion and Conclusion: Created erosion resulting of this ROAD NETWORK was estimated about 1576.4 ton/year without considering SEDMODEL. This amount was estimated about 140.7 ton/year with considering SEDMODEL on basis distance of waterway. Results showed that the share of ROADs sediment yield is about 1.33 percent of watershed exit, but ROADs includes 0.017 percent of watershed areas. Pay attention to covered areas by ROADs in watershed, the amounts of sediment yield is 78 equal of sediment yield average in watershed area.

Regarding to the importance of ROAD NETWORK in FOREST Management, Wood transportation and FOREST operation, this investigation was carried out in four districts of "Vaz" experimental FOREST. The study area was 4081 hectares. The FOREST NETWORK planning was based on essential factors such as length of ROADs, length density of ROADs, wood transport distance in FOREST, NETWORK percentage, slope, soil mechanics, soil stability, volume of wood production and climatic factors. Three stages were considered to achieve these goals: In the first stage, existing ROADs were evaluated and recorded. The results were as follow: Length of ROADs: 21770 m, length density of ROADs: 7.8 m/ha and NETWORK percentage: 37 Due to unacceptable of the existing NETWORK, in the second stage FOREST NETWORK planning (Major ROADs and Minor ROADs) was performed based on the important factors mentioned above to obtain optimal ROAD coverage and to reduce FOREST damages. In the third stage, new ROAD NETWORK was planned for each district and marked on its natural site, based on the factors mentioned above. The new ROAD NETWORK characteristics were as follow: Length of ROADs: 52469 m, length density of ROADs: 12.8 m/ha and percentage of NETWORK: 60 The soil mechanics test showed that soil is generally heavy textured with low plastic limit and high liquid limit. The result showed that although ROAD length density was increased but percentage of ROAD NETWORK was not expanded, because steep areas of the FOREST was a limiting factor for major ROAD NETWORK. For this reason, secondary ROAD NETWORK was designed to complete the total ROAD cover. As a result, the total length and density of the planned ROADs were 74459m and 19.5 m/ha, respectively. The density and percentage of minor ROAD NETWORK in series 1,2 and3 were appropriate and covered a desirable area of the FOREST. Slope was an important factor in major and minor ROAD planning. Only 43% of the ROADs were located on the best class of slope (O-30%).The logging ROADs were located on 62% of the FOREST ground with moderate mechanical class of soil. Soil stability and climatic factors had important roles in logging ROAD planning. In district 2, the slope was suitable but the soil stability was a limiting factorin logging ROAD construction all over its area. As a result, cable crane system should be applied to transport logs outside the FOREST. The soil mechanics test showed that soil is generally heavy textured with low plastic limit and high liquid limit.