FLOOD FORECASTING SYSTEM UPGRADE BY ADDING ICE PHENOMENA

Janina Brastovickytė-Stankevič1

1 Laboratory of Climate and Water Research, Geology and Geography Institute, Nature research centre, Lithuania

Modelling natural process is a very complex matter affected by many factors. In this study for water level forecast Mike Customised (MC) forecasting system is used. To get a good simulations, everyday on scheduled time system should get a new information from weather radar, numerical weather forecasting system, water gauging stations; and meteorological stations. After processing received data flood forecasting system can make water level forecast simulations.

Winter of 2015-2016 was the first one when the MC system was started to run. Forecasting system using all received data calculates and simulates some soil, snow and other parameters which are used to simulate reality and to make water level forecast. In this study, a decision was made to try and improve forecasting system by adding actual ice phenomena. Ice phenomena as a parameter is extremely difficult to value. Especially ice jams, that determines the sudden water level changes.

To understand the need of entering actual ice phenomena into the forecasting system, a few main characteristics has been chosen: ice coverage, slush/debacle, ice jam and ice (for other ice phenomena). For these parameters Manning roughness coefficient from 0,13 to 0,4 was assigned. For ice phenomena effect on water level forecast these water gauging stations have been selected: Nevėžis – Babtai, Neris – Jonava, Nemunas – Nemajūnai, Nemunas – Panemunė, Merkys – Puvočiai and Jūra - Tauragė. In these stations simulations with and without entered ice phenomena was compared.

The Study showed that entering ice phenomena didn't help to improve forecast, it makes forecast less accurate. Model is calibrated to give better forecasts, and for more accurate results Manning roughness coefficient should be changed.