Modeling landscape hydrological response to spatial–temporal land use/cover changes

The north of Iran is facing high flood frequency and economic and natural losses even though it is the greenest in the country. In order to understand the functional changes, the hydrological response to spatial–temporal land cover/land use (LCLU) changes was investigated in Gharesou Watershed of Golestan Province, Iran. For this, structural changes in LCLU were derived using landscape metrics of fragmentation for the years 1984, 1994, 2014 and 2030. For quantification of average annual runoff as hydrological function, we utilized the Long-Term Hydrologic Impact Assessment (L-THIA) model. The annual average runoff production in the whole watershed showed an increase during 1984 to 2014 from 6.15 to 6.78 m3/0.09 ha/yr and was predicted to continue, reaching above 6.98 m3/0.09 ha/yr in the year 2030 due to more fragmentation and degradation in forest cover. The results of statistical analysis between structure and function highlighted four behavior clusters in LCLU types and the generated runoff (F (6, 20) = 273.36, p = 0.000). The forest, grass lands and water bodies behaved similarly in a cluster with the lowest runoff production, whereas industrial and high-density residential LCLU were responsible for the highest runoff generation. Linear regression indicated that the metric of mean nearest-neighbor distance was superior in predicting the runoff volume. As control of the potential runoff is an important factor in future development of LCLU, the connectivity of natural areas should be conserved and that of man-made land uses should be reduced in future.