Application of Classification Tree Analysis (CTA) to Model the Potential Distribution of Harmful Algal Blooms (HAB’s)

Abstract

Remote sensing has a potential inconventional approach to observe the quality of lake as well as other waters land. Riam Kanan is a reservoir which has a water resource from Riam Kanan River with the wide of watershed is 1043 km. The accumulation of nutrient simultaneously causes the condition of waters at reservoir is getting thriven. The thriven water condition can cause an increasingly growth of harm micro algae or Harmful Algal Blooms (HABs). This research tries to apply Classification Tree Analysis (CTA) method to model the potential distribution of HABs which uses image of satellite Landsat-8 OLI. Landsat 8 OLI image which was recorded on 14 August 2016 was used in this research based on value at surface reflectance. Classification Tree Analysis (CTA) method was used to model the potential distribution of HABs at Riam Kanan Reservoir. The result of CTA model then was used to analyse the parameter that affect the potential distribution of HABs. Based on the result of modelling with the total validation model 81,25 %, it is resulted that there are 4 potential classes, they are light, medium, heavy, and extremely heavy classes which the distribution of HABs in a high depth is dominated by medium class, whereas in shallower depth with area of waters that stick out into the water is included in heavy potential class. Potential of load pollution is obtained from outer part of the reservoir especially from dry land agriculture in width 303.811,95 Ha which is known has the amount of potential Nitrogen content of 20.507.306,6 kg and phosphorus with the total content as much as 4.557.179,25 kg.