SALLJEX RAINFALL ANALYSIS

Peruvian Altiplano Rainfall Analysis

The dataset obtained from the Altiplano raingauge network revealed the presence of the region of largest rainfall over Lake Titicaca. Figure 1 shows an analysis performed using 98 SALLJEX raingauges, which were selected through quality control procedures. The period chosen for the analysis is Dec.1.2002 - Feb.28.2003 since most of the SALLJEX temporary networks operated continuously during this period resulting in a quite continuous dataset to work with.

The largest rainfall observation was 636 mm observed in Okawasi (in Isla Soto), which is located in the northern part of the lake. A region of large rainfall in this part of the lake has been suggested by analyses sketched by different authors in the past. The 636 mm accumulated in Okawasi correspond to 212 mm/month. Rainfall over 500 mm was observed in other locations among Lake Titicaca. Some of them are Moho (541 mm), Conima (549 mm) and Patascachi (515 mm) to the northeast; Lampayuni (534 mm) and Occosuyo (514 mm) in the island of Amantani located to the west; Huayllata (501 mm) in the Ilave Peninsula located to the west; Unicachi (555 mm) in the Copacabana Peninsula located to the south. This large rainfall quantities are associated to the lake-effect nocturnal convective storms that originate over the lake. This explains the decrease of the 3-month rainfall accumulations away from the lake axis.

Figure 1. Rainfall (mm) accumulated from December 1, 2002 through February 28, 2003 in the Peruvian Altiplano. The data utilized for the analysis is daily data from 98 SALLJEX & SENAMHI daily-reporting raingauges selected through quality control procedures among a network of 166 stations. Lake Titicaca is indicated with a thick line and horizontal lines as a background. The terrain located above 4500 mm has been shaded with light gray.

A secondary region of high rainfall was detected in the vicinity of the high terrain located to the west of Lake Titicaca. 3-month accumulations as high as 589 mm were observed. This maximum, however, was produced by afternoon convective storms. These have a larger signal on the IR4 satellite imagery when compared to the lake effect storms, but they generate about as much rainfall or less than the nocturnal clouds that form over the lake.

A third maximum appears in the northeastern section of the figure, but it is located outside of the Altiplano. This region, located in the eastern slopes of the Andes, is associated to convection produced by the upslope flow east of the Altiplano.

Less rainfall can be observed over relatively flat terrain away from Lake Titicaca, in particular to the south and southwest were the availability of low level moisture is reduced compared to the north and northeast. Rainfall rates as low as 178 and 184 mm (~ 60 mm/month) are observed in Viluyo and Laraqueri respectively.

Rainfall Timeseries and the effects of Lake Titicaca

Figure 2 shows the timeseries of rainfall calculated by averaging the information of the 98 sites selected. The blue line shows this average and helps to visualize the rainy and wet spells in the Peruvian Altiplano during the SALLJEX. The orange line represents the timeseries of rainfall calculated by averaging the data from 6 raingauges located in the islands of Taquile, Amantani and Soto. This line indicates the variability of the lake effect storms. The red dotted line indicates the number of raingauges utilized for the all-sites-average, which is equal to the number of raingauges reporting data.

Only about 20 raingauges from the 98 selected were operating suring October. They correpond to the Peruvian National Weather Service's (SENAMHI) permanent network. The installation campaign in the northern sector of the Altiplano started by mid October which explains increase in the nmber of observations between October 20 and November 5. The network correspondent to the southern sector of the Altiplano was installed between November 10 and November 20, 2002.

Figure 2. 98-site-averaged daily rainfall timeseries from October 1, 2002 through May 31, 2003 in the Peruvian Altiplano. The continuous blue line represents the daily rainfall averaged over all the reporting raingauges. The number of raingauges used for the average is indicated with a dotted red line.

The average obtained using the 98 sites represents a fair index of dry and wet spells in the Altiplano, in particular in the vicinity of Lake Titicaca, where the densest network was deployed. This figure shows that rainfall occurs in the form of wet and dry spells of several days of lenght, as described by other authors in the past. The longest and most intense wet spells occurred in late January and in late February through early March, being the January spell the most intense and prolongued. Averaged rainfall rates as large as 14 mm/day per station occurred.

The six stations located among the islands showed larger day to day variations both due to a smaller sample size and due to larger rainfall rates over the lake when compared to the entire raingauge network region. This timeseries suggests that lake effect storms occur mainly during rainy periods in the Altiplano. These storms, however, are not produced every rainy day in the Altiplano as a higher frequency oscillation is visible in the figure. Interested in this variability we extracted the period December 1, 2002 - February 28, 2003 and calculated the difference between the lake average and the 98-site-average (LAKE-ALL). The 30 most extreme values were selected and dated in figure 3. The dots indicate 15 strongest lake-effect-storm days, and the triangles the 15 weakest lake-effect-storm days.

Figure 3. Difference between the 6-lake-stations rainfall average and the 98-stations rainfall average. The blue dots indicate the 15 days in which the rainfall over the lake was the largest compared to the overall rainfall field. The triangles indicate the 15 days in which the average rainfall exceded the rainfall reported over tha lake by the largest amounts.