Study On Efficiency Improvement Of Train Dispatch Under Severe Rainfall By Using Radar

---

Friday 3 july 2015

13:00 - 13:03h at Europe 2 (level 0)

Themes: (T) Extreme events, natural variability and climate change, (ST) Flood resilient cities and infrastructures, Poster pitches

Parallel session: Poster pitches: 15I. Extreme - Society & Resilienty

---

Recently, the frequency of severe rainfall increases in Japan. The average of annual frequency of severe rainfall which is defined as rainfall having more than 50mm of rainfall within one hour has become increase from 168 times per year (TPY) during the year of 1976-1986 to 226 TPY, 1999-2010. The increasing of frequency of severe rain raise risk of sediment related disaster and inundation disaster. Trains especially local train run on reclaimed land and mountainous area have risk of such disasters when severe rain might happen. Therefore, railway companies take many kinds of countermeasures. The present study deals with the dispatching problem under severe rainfall. Each train dispatching system has defined a threshold of rainfall amount where dispatchers stop trains or they command to reduce the speed of trains. A Japanese railway company has measured the rainfall amount by their own rain gauges that locate almost every 13km and it is very difficult to change the measurement system because the company has a historical rainfall data by which the threshold of rainfall amount has been defined. On the other hand, the company has an interest for introducing radar rainfall information for dispatching trains because of increasing of severe rainfall having small spatial scale and efficiency improvement for collecting rainfall data. Many railway companies have the similar contradiction. The present study shows a statistical analysis of the uncertainty to use radar for capturing severe rainfall by using probability function of difference between rainfall amount measured by rain gauge and one by radar.