Landslides and vegetation cover in the 2005 North Pakistan earthquake: a GIS and statistical quantitative approach

P Peduzzi

doi:10.5194/nhess-10-623-2010

Abstract

Abstract. The growing concern for loss of services once provided by natural ecosystems is getting increasing attention. However, the accelerating rate of natural resources destruction calls for rapid and global action. With often very limited budgets, environmental agencies and NGOs need cost-efficient ways to quickly convince decision-makers that sound management of natural resources can help to protect human lives and their welfare. The methodology described in this paper, is based on geospatial and statistical analysis, involving simple Geographical Information System (GIS) and remote sensing algorithms. It is based on free or very low-cost data. It aims to scientifically assess the potential role of vegetation in mitigating landslides triggered by earthquakes by normalising for other factors such as slopes and distance from active fault. The methodology was applied to the 2005 North Pakistan/India earthquake which generated a large number of victims and hundreds of landslides. The study shows that if slopes and proximity from active fault are the main susceptibility factors for post landslides triggered by earthquakes in this area, the results clearly revealed that areas covered by denser vegetation suffered less and smaller landslides than areas with thinner (or devoid of) vegetation cover. Short distance from roads/trails and rivers also proved to be pertinent factors in increasing landslides susceptibility. This project is a component of a wider initiative involving the Global Resource Information Database Europe from the United Nations Environment Programme, the International Union for Conservation of Nature, the Institute of Geomatics and Risk Analysis from the University of Lausanne and the "institut universitaire d'études du développement" from the University of Geneva.

Highlights

Overexploitation of natural resources and deforestation is one of the main triggers for the observed increase in landslide disasters along with increase in population exposure (Nadim et al, 2006)
This paper describes how spatial and statistical analysis using remote sensing and Geographical Information System (GIS) techniques were applied
The study confirmed the hypothesis that landslide occurrence is higher on steep slopes, close to rivers, trails, active fault and that vegetation cover seems to act as stabiliser in this region

Summary

Introduction

Overexploitation of natural resources and deforestation is one of the main triggers for the observed increase in landslide disasters along with increase in population exposure (Nadim et al, 2006). The Hyogo Framework for Action (HFA) “encourages the sustainable use and management of ecosystems, [for] reducing the underlying risk” (UNISDR, 2005). To achieve this goal, both local authorities and international decisionmakers need to adopt improved environmental policies. There is a need for a multiplication of scientific evidence and for developing simple methods allowing solid scientific assessments Outputs from this quantitative analysis were used in an interdisciplinary study for disaster risk reduction (SudmeierRieux et al, 2008). It explored the relation between land use factors, such as deforestation, grazing, road building, etc. It was applied and tested on the area affected by the earthquake that hit North Pakistan

Objectives

Methods

Findings

Discussion

Conclusion