Climate Change Impacts and Adaptation National Conference
Conference Home Page
Conference Venue
Conference Program
Field Trips
Accommodation for conference
Abstracts-Papers
Abstracts-Posters
Link to Registration site
Conference Contact
Français

Abstracts - Posters

Long-term Predictions of the Red River Floods (RRF)

Boris Berri (Berry)
Health Canada
bberri@sympatico.ca

The anomalous variability of the meteorological conditions has a planetary character and activates dangerous natural phenomena. The extremal points (reference years) of the harmonic Model of Northern Hemisphere Temperature Anomalies (MNHTA) connect with dangerous anomalous dispersions of the climate-forming factors (Berry at al., 1986). The MNHTA was detected from the tree-ring series (1660-1965). Predictions of the southern oscillations El NiZo (ENSO) and activations of snow avalanches with accuracy +/-1 year already for 40 years were justified with a probability about 70%.

The pattern of the model's reference years was moved forward for one year for the better coincidence of the model's and RRF's years. The reference years, which have significant Amplitudes (Am>0,0685 °C), calculated from 31-year running mean zero line, coincide (the standard deviation +/-1.1y) with the years of the 13 RRF with significant Discharges (D>2000m3/s) for the adjusted time interval (1850 -2000). The Am and crucial D have strong positive correlation (r = 0.917, the Level of Significance, LS< 0.001): D=1.017+11.325*Am.

The analysis of variance showed there are real possibilities of the RRF' long-term forecast both in relation of the times' prediction with precision +/- 1.1y and the discharges' prediction with accuracy (10+ / - 3)%. The probability of significant floods (D>2000m3/s) for 1998-2003 was <0.7% and for 2004 was about 10%. The next flood will take place with a probability of about 55% in 2006+/-1. The predicted D = 2560 m3/s+/-10% at Redwood Bridge (Winnipeg) is similar to the flood of 1966.


2005-04-06

top of the page