By Ujang Wisnu Barata – (Thesis M.Si Sriwijaya University, 2013). Supervised by Prof. Charlotte de Fraiture Ph.D., M.Sc. (UNESCO-IHE), Prof. Ir. Robiyanto H. Susanto, M.Agr.Sc., Ph.D (Sriwijaya University), F.X.Suryadi, Ph.D., M.Sc. (UNESCO-IHE).

Gunung Leuser National Park (GLNP) area in a section of Langkat region have been experiencing severe degradation and deforestation as a result of illegal logging and encroachment for land use conversion. These facts resulted in critical condition of some watersheds that tipped in GLNP. Several times the incidence of flooding, with a peak incidence in last 2006, gives the fact that causal relationship between environmental degradation and disasters are not indisputable.

One of the areas that have severe degradation and deforestation is the upstream part of Sub-watershed Besitang. According to the results of GIS analysis using landsat imagery in 2009, the number of degraded and deforested area is amounted to 7,168 ha or 19% of the total catchment area inside the national park. The scope of this study focused on flooding event in the most prone rural areas in Besitang District, which are Bukit Mas Village; Sekoci Village; and PIR ADB Village.

DUFLOW Model was used to construct and to predict a river flood model based on three scenarios of upstream condition. Scenario 1 was assumed as the ideal condition of GLNP which is represented by the intact forest formation that can fully handle the flood prevention, with 3% of runoff coefficient; Scenario 2 described the existing condition in which some disturbed areas were occurred, with 25% of runoff coefficient; and Scenario 3 was assumed to be the worst condition due to land conversion into non conservation use that affected to more impervious cover in GLNP, with 95% of runoff coefficient. Hence GIS were used to analyze the type of land use which is affected by the inundation and to produce flood risk map. Finally, the substitute cost analysis was obtained to compare the flood prevention value of GLNP with the damage due to flooding.

The study resulted that the land use change in the national park is the main factor that can accelerate flooding. The ideal condition which is represented by Scenario 1 resulted no flooding. This condition indicates the important role of intact forest formation in handling the runoff. On the other hand, Scenario 2 and Scenario 3 have shown increasing inundation depth, area of inundation, and flooding duration due to more impervious cover in the upstream which can lead more runoff. The flooding characteristics at Scenario 2 and Scenario 3 were:

• Scenario 2, with 2 years of return period; total inundated areas are 31.76 ha, with the inundation depth not more than 2 m, and 23 hours of flooding duration;
• Scenario 2, with 10 years of return period; total inundated areas are 65.72 ha, with the inundation depth > 2 m covered 8.67 ha, and 30 hours of flooding duration;
• Scenario 2, with 50 years of return period; total inundated areas are 88.91 ha, with the inundation depth > 2 m covered 23.36 ha, and 32 hours of flooding duration;
• Scenario 3, with 2 years of return period; total inundated areas are 342.02 ha, with the inundation depth > 2 m covered 186.03 ha, and 40 hours of flooding duration;
• Scenario 3, with 10 years of return period; total inundated areas are 450.79 ha, with the inundation depth > 2 m covered 340.39 ha, and 42 hours of flooding duration;
• Scenario 3, with 50 years of return period; total inundated areas are 508.41 ha, with the inundation depth > 2 m covered 421.86 ha, and 43 hours of flooding duration;

In Scenario 3, which illustrated the worst condition, at 50 years of return period, the damage cost in the three villages is more than 1 million € within 508 ha total inundated land uses. The damage cost evaluation which was done by determining type of inundated land use described the scale of economic and environmental damage. The type of inundated land use and properties in which their economic values can be calculated were plantation areas and its commodities, agriculture areas and its commodities, number of houses, and public facilities. The environmental damage have described by calculating the reduction of land price per hectares. The cost benefit analysis also computed the flood prevention value of 37,000 ha upland areas in GLNP by determining the reforestation cost compare to levees construction. It resulted 7.7 million € for ecosystem restoration, and 9.5 million € for levees construction.

Finally, structural and non structural measures have to be taken to mitigate flooding which are dyke constructions as structural measures, and also ecosystem restoration of GLNP, land use evaluation, and awareness enhancement by considering the government regulation as non structural measures. However, the cost analysis shows that conducting ecosystem restoration is more feasible than constructing dyke.

The management authority of Gunung Leuser National Park should initiate an integrated program on ecosystem restoration and flood risk management. Thus, the community, especially in the upland, need to be empowered as a strategic partner to preserve the forest and actively participate in the recovery effort. All key stakeholders i.e. government entities, limited companies, communities, NGOs, and other relevant agencies should be actively involved to re-evaluate the land use to cope with extreme flood event by implementing the Government Regulation No. 38/2011 on River Management and No 37/2012 on watershed Management. The calculation of flood damage due to flooding would likely result greater value if the scope of study area are extended. Referring to the flood event in 2006, the downstream areas which are more densely populated with numerous property assets were affected severe enough. However, the DUFLOW model could not describe the flood magnitude in the downstream areas due to unavailability of river dimension / cross section and also tidal fluctuation data. Therefore the measurements of canal dimension and regular recording of tidal fluctuation are need to be done to get a comprehensive overview of flooding impact in a whole watershed areas.

Keywords: Land use, Gunung Leuser National Park, flood risk, DUFLOW model, GIS, flood damage