[p1] Applications in various physical and biological systems


An OSSE of the response to Typhoon Man-Yi on Lake Biwa, Japan

G. Auger (IBM), J. Wells (Ritsumeikan University)


For disaster mitigation purposes, notably tracking of pollutan, we have been developing a Nowcasting system for Lake Biwa, to the east of Kyoto, with simulations performed by the SUNTANS ocean model. In this work, we perform on OSSE with meteorological forcing from around the passage of Typhoon Man-Yi on September 16th 2013. Assimilation is performed, using synthetic temperature measurements from the locations of 10 operational monitoring gages, by the Local Ensemble Transform Kalman Filter, with Bred Vectors applied to generate an ensemble of estimates of the state vector. We used 6 perturbed runs, adding uniformly distributed noise in the range [-0.2 0.2] to the temperature field, and 1 control run. We used the Root Mean Squared Fluctuation of an ensemble's temperature distribution to rescale the bred vectors every 6 hours. Identical wind histories, namely those interpolated from meteorological measurements, were applied to all runs in the ensemble. With this important caveat, the ensemble-mean forecast tracks the nature run satisfactorily over three weeks in terms of bias and RMS error of temperature, and surprisingly well in terms of the spatial distribution of surface temperature. An offset of about 1.2 degrees that was applied to the forecast at the beginning decreased to about 0.4 degrees within 48 hours, and continued decreasing thereafter.