Download full issue - PanamJAS

Climatic variations in Southern Brazil255excessive discharge is associated with largeinterannual variations of rainfall over the basin. Theaverage precipitation rate over the Patos-Mirimdrainage basin is 1,200 mm/year. Above averagerainfall in excess of 2,000 mm/year was observedduring El Niño events and the lowest average valueon record (near 800 mm/year) was observed duringthe 1988 La Niña (LN) (Beltrame & Tucci 1998).The mean surface area of ML is approximately3,749 km 2 (185 km long and 20 km wide). Itscatchment basin includes almost 55,000 km 2 (47% inBrazil, 53% in Uruguay) and ML is linked to thePatos Lagoon estuary through a 76-km narrownatural channel called São Gonçalo.The term “coastal lagoon” usually refers towater bodies along the coast with one or moreconnections with the ocean (Bird, 2008). Therefore,although ML is traditionally known as a coastallagoon, it resembles more closely an overflowinglake. In the past, Patos Lagoon brackish waterscould reach ML through the São Gonçalo channel,damaging inundated rice crops in the region. Toavoid salt penetration upstream, a subsurface damwas built across the channel in 1977 to block denserbrackish waters but still allowing surface fluxes andnavigation. The dam is 3.2 m high and is placed in across section where the mean depth is around 5 m.After the dam was built, ML waters remained freshall the time, with an estimated mean overflow intoPatos Lagoon of 700 m 3 /s (Machado 2007). Changesof outlet conditions (as the subsurface dam) shouldonly marginally influence mean overflowing lakelevels (Bengtsson & Malm 1997).In addition to rainfall-runoff relationship,the residence time of Patos-Mirim system is alsoinfluenced by synoptic scale atmosphericphenomena. With a narrow connection restrictingfresh-seawater exchanges, circulation within thesystem is driven by the combined effect of windsand runoff.Dynamically, the passage of a cold front andthe associated strong south-southwesterly windsdrive a water level set up in the northern region ofboth Mirim and Patos Lagoon (e.g. Möller et al.2001). This could favor ML discharge, but alsodrives shelf waters against the coast, pushingseawater into the estuary and balancing the pressuregradient. Thus, seawater inflow and the wide floodplains along São Gonçalo channel slow down MLrunoff. With the weakening of the southerlycomponent of the winds and the establishment ofnortheasterlies, seawater retreats and the southwardpointingpressure gradient inside the lagoons acts toequalize the water level throughout the system,driving Patos Lagoon waters towards its estuarinearea. This effect, together with the southwardmovement of ML waters, partially dams ML again,increasing its residence time. Northeasterly windsare predominant during the whole year withincreasing importance of southwesterlies duringwintertime, associated with atmospheric cold frontspropagating over the region on time scales rangingfrom 3 to 11 days (Stech & Lorenzetti 1992).In general, Southeastern South America(SSA - Southern Brazil, Northeastern Argentina andUruguay) experiences positive rainfall anomaliesduring EN events and negative anomalies during LNevents (Grimm et al. 1998, 2000). This associationalso holds for streamflow anomalies, which presentinterannual cycles of 3.5 and 6.3 years, coherentwith El Niño – Southern Oscillation (ENSO) cycles(Robertson & Mechoso 1998).The relationship between ENSO and riverrunoff in Negro and Uruguay Rivers were exploredby Mechoso & Perez-Irribaren (1992). They found atendency for below average streamflow from Junethrough December during LN events and a slightlytendency for above average streamflow fromNovember through February during EN years.Figure 1. Map showing the La Plata River basin with(1 - Paraguay River, 2 - Paraná River, 3 - Uruguay Riverand 4 - Negro River) and Patos-Mirim basin. The detailshows the locations where the water level is observed (a -Santa Isabel and b - Santa Vitória do Palmar). The thickblack line represents the São Gonçalo channel.Pan-American Journal of Aquatic Sciences (2010), 5(2): 254-266