88th Annual Meeting (20-24 January 2008)

Wednesday, 23 January 2008
Study of the frequency of extreme daily precipitation on the south of Andes mountain range . Temporary variability in the period 1961-2003 and relation with the Antartic Oscillation
Exhibit Hall B (Ernest N. Morial Convention Center)
Federico Ariel Robledo, Departament of Atmospheric and Ocean Sciences, Universidad de Buenos Aires, Buenos Aires, Argentina; and O. C. Penalba
Poster PDF (136.9 kB)
On the center of the Patagonic mountain range, on the provinces of Chubut, Neuquén and Rio Negro, Southern Argentina, seven hydroelectric dams are located, in a relatively small region compared to the surface of Argentina. The energy production, in an average annual term, that these hydroelectric dams have produced is of 15741 Gwh. This value represents, approximately, 16.7% of the total energy that has been produced in Argentina in 2004. The water of the melting of snow or from glacier of the mountain range is the main source for electrical generation. In addition, these dams are used to regulate the flow of rivers, activities of irrigation and industries or human consumption.

There are no records of snowfalls over the mountain range in this region of Argentina and only few raingauges in patagonia region. For this study were used daily precipitation of Esquel (43º S - 71ºW), located at the east of the Andes mountain range. This station has long record (1961-2003) without missing data.

The objectives of this research are to provide a study of the temporal variations of different extreme climate indices, and make a first approximation of the relation between these indices and the circulation in the South Hemisphere. Different indices are analysed: frequency of daily precipitation, considering different threshold; persistence of rainday and how much of the yearly index comes from each season index and a quantification of these changes. For the second objective, the Antartic Oscillation Index is analysed as an estimator of the circulation in middle and high latitudes.

In order to evaluate rainday the 0.1mm threshold is considered and daily extreme precipitation has been calculated using different thresholds by means of 50th, 75th, 90th and 95th percentiles (thereafter P50, P75, P90 and P95, respectively). These percentiles are calculated for every day of the year in period 1961-2003. For the three superior percentiles, an annual cycle was observed, with greater values in winter with respect to the summer. For example, P75 shows a maximum in winter of 20mm/day and a minimum in summer: 7mm/day.

The 33% of the winter days present precipitations over 0,1mm, whereas in summer is reduced to 13%, for the climatology corresponding to the period 1961-2003. The percentage of days (PE) of extreme rain from the P75 shows an annual cycle of precipitation: in average, 8% of the winter days registers extreme precipitation, whereas in summer the value is reduced to an average of 3%. The Extreme daily precipitation intensity, shows an annual cycle different from the percentage of events. In winter, 21mm/day are registered, while during Spring and Autumn the intensity is of 11mm/day. During Summer the intensity is of 16mm/day. The probability that it rains, when the previous day was rainy (persistence of the rainy day (P11)), also shows an annual cycle. Winter probability its around 0.55 and in summer 0.35. Spring and autumn present values between 0,40 and 0,45. Winter explains 36% of the annual daily precipitation, both for the threshold 0.1mm and for the given by the P75 percentile. However, summer explains 15% of daily rain. Autumn and Spring explains 27% and 22% respectively, of annual daily precipitations.

The tendency of the temporal variability of the percentage of rainday and the P75 was calculated. In order to evaluate tendencies used the non-parametric method of Kendall Tau. In winter, for both thresholds, negative and non significant tendencies was observed, in the period 1961-2003.

The percentage of change of the amount of rainday and the extreme events between periods 1961-1975 and 1980-1996, was calculated for Summer, Autumn, Winter and Spring. Winter shows a 36% less of precipitation events over p75 in the second period (1980-1996). Summer and spring also show a decrease (19% and 14% respectively). In annual terms a decrease of 18% is observed. For the percentage of change of rain over the threshold 0.1mm the decrease is not so clear. Winter shows a 6% less of days of rain. However Autumn and Spring show an increase of 12% in average. In annual terms an increase of 4% of the days of rain in the second period is observed.

The Antartic Oscillation Index (AAO) was used, as an estimator of the circulation of the South hemisphere in middle and high latitudes. The AAO index obtained from the Climate Prediction Center from data of reanalisis NCEP. Correlation between the monthly value of AAO and the monthly anomaly of percentage of extreme precipitation calculated, shows a value of -0,14; which is significant statistically to 5%, in the period December 1979 to January 2003. It has been observed a positive anomaly of amount of daily precipitation extreme, during the negative phase of the AAO. However, it has been observed a negative anomaly of amount daily precipitation extreme precipitation, during the positive phase of the AAO. During Autumn, correlation is -0,54 which is significant to 1% and -0,27 in Spring it is significant to 10%. In Summer and Winter the correlation is negative but nonsignificant. The phase of the AAO in autumn explains a 29% of the amount of extreme daily precipitation, whereas in Spring it explains a 7,29%.

Supplementary URL: