OPTIMIZING FARM-ROAD ACCESS IN AGRICULTURAL LAND-USE PLANS: A GIS NETWORK EXPERIMENT FOR SAMARKAND REGION

Abstract

Farm-road design must distinguish road proximity, direct field access and travel through a connected network. This study expands a reproducible 100-ha GIS experiment intended to inform land-use planning in Samarkand Region. One 10 × 10 field layer is evaluated under a perimeter baseline, a central cross, two parallel internal roads and a full-coverage alternative. Exhaustive enumeration of 4,048 subsets of 18 full-length candidates gives maximum coverage of 36, 52, 68, 84 and 100 fields for zero to four additional kilometres. At the same 2-km budget, the central cross serves 64 fields and the parallel alternative 68, although the cross gives a slightly smaller mean centroid-to-road distance (90 versus 94 m). On the same 36 initially served fields, full coverage reduces mean network distance from 997.2 to 986.1 m, or 1.11%. A 4-m corridor occupies 1.60 ha in the full-coverage alternative; doubling its width doubles the footprint without changing the centre-line access model. Hypothetical cost–benefit scenarios include both positive and negative outcomes. Five computed model maps, alongside the supplied regional location map, separate proximity, access gaps, routes and network design. The results are deterministic benchmark findings, not observed road conditions or approved projects in Pastdarg‘om. They show why coverage, matched routes, land take and feasibility must be reported separately.