Tag Archives: Correlation techniques

Using Correlative Data Analysis to Develop Weather Index That Estimates the Risk of Forest Fires In Lebanon: Assessment versus Prevalent Meteorological Indices (Published)

Forest fires are among the most dangerous natural threats that bring calamities to a community and can turn it totally upside down. Lebanon is considered one of the countries that face this natural disaster especially in summer season. Prevention is considered as one of the very essential tools to cope with and overcome such a danger.  This is especially true in developing countries where fire suppression cannot be affordable.  Early warning fire danger rating systems have been adopted by many developed countries to decrease fire occurrence. In this paper, data analysis is used to find the most affecting parameters on fire ignition during the last six years in north Lebanon using different correlation techniques: statistical regression, Pearson, Spearman and Kendall’s Tau correlation. The correlations of these attributes with fire occurrence are studied in order to develop a new fire danger index. The strongly correlated attributes are then derived. The index is a simple linear equation relying on few numbers of weather parameters that are easy to measure and which facilitate its application in developing countries like Lebanon. The outcomes resulting from validation tests of the proposed index show high performance in the Lebanese regions. It is strongly believed that this index will help improve the ability of fire prevention measures in the Mediterranean basin area.

Keywords: Correlation techniques, Data analysis, Fire danger Index, Forest Fire prediction

Using Correlative Data Analysis to Develop Weather Index That Estimates the Risk of Forest Fires in Lebanon: Assessment versus Prevalent Meteorological Indices (Published)

Forest fires are among the most dangerous natural threats that bring calamities to a community and can turn it totally upside down. Lebanon is considered one of the countries that face this natural disaster especially in summer season. Prevention is considered as one of the very essential tools to cope with and overcome such a danger.  This is especially true in developing countries where fire suppression cannot be affordable.  Early warning fire danger rating systems have been adopted by many developed countries to decrease fire occurrence. In this paper, data analysis is used to find the most affecting parameters on fire ignition during the last six years in north Lebanon using different correlation techniques: statistical regression, Pearson, Spearman and Kendall’s Tau correlation. The correlations of these attributes with fire occurrence are studied in order to develop a new fire danger index. The strongly correlated attributes are then derived. The index is a simple linear equation relying on few numbers of weather parameters that are easy to measure and which facilitate its application in developing countries like Lebanon. The outcomes resulting from validation tests of the proposed index show high performance in the Lebanese regions. It is strongly believed that this index will help improve the ability of fire prevention measures in the Mediterranean basin area.

Keywords: Correlation techniques, Data analysis, Fire danger Index, Forest Fire prediction