|
Every year, Jakarta, the capital of Indonesia, always experiences flooding. When natural disasters such as floods occur, the government is obliged to carry out a series of disaster emergency responses, one of which is disaster modeling. Disaster modeling can be done with various data sources, but field survey data sources, remote sensing, and aerial photographs are considered less efficient to use. Field surveys require a longer time, while remote sensing and aerial photography have limitations during the rainy season, which is disrupted due to high cloud intensity. Social media, especially Twitter, has currently received a lot of attention from various groups as a source of data for flood modeling. This research aims to build a spatial database, perform flood modeling, test the level of accuracy produced, and test tweet data with rain events in Jakarta. Flood modeling is carried out with the Kernel-Based Flood Mapping Model method using DEMNAS data, river distribution data, administrative boundaries, water depth data at each river floodgate, tweet data, BPBD flood area maps, and surface observation rainfall data issued by BMKG. The results of flood modeling were tested for accuracy with the overall accuracy method against all tweet data obtained and BPBD flood area maps. In addition, a regression test was conducted to determine the relationship between rainfall and tweet data related to flooding in Jakarta. The results showed that 149 out of 12,345 tweets could be compiled into a database that was used as the basis for modeling. Flood modeling results show an accuracy value of 70% based on the calculation of total flood points and 57% based on BPBD flood zones, this value is in a low category. The regression test between flood points and January rainfall data shows a relationship that does not affect. The regression test value between rainfall at the flood location and the number of tweets is 0.020822, while the flood depth is 0.049214, which means that the rainfall variable only affects the variable number of tweets by 2% and the depth of the flood by 4.9% and the rest is influenced by other factors.
|
