Vargas tragedy

Vargas tragedy
A part of Vargas state after the 1999 mudslides
Date14 December 1999 (1999-12-14) – 16 December 1999 (1999-12-16)
LocationVargas State, Venezuela
Coordinates10°36′18.67″N 66°50′58.21″W / 10°36′18.67″N 66°50′58.21″W / 10.6051861; -66.8495028
Vargas in Venezuela (special marker).svg
Location of Vargas in Venezuela

The Vargas tragedy was a natural disaster that occurred in Vargas State, Venezuela on 14–16 December 1999, when torrential rains caused flash floods and debris flows that killed tens of thousands of people, destroyed thousands of homes, and led to the complete collapse of the state's infrastructure. According to relief workers, the neighborhood of Los Corales was buried under 3 metres (9.8 ft) of mud and a high percentage of homes were simply swept into the ocean. Entire towns including Cerro Grande and Carmen de Uria completely disappeared. As much as 10% of the population of Vargas died during the event.[2]


A section of Los Corales, one of the neighborhoods in the Vargas state that suffered the heaviest destruction

The coastal area of Vargas State has long been subject to mudslides and flooding. Deposits preserved on the alluvial fan deltas here show that geologically similar catastrophes have occurred with regularity since prehistoric times.[2] Since the 17th century, at least two large-magnitude debris flow, landslide, or flood events, on average, have occurred each century within the modern boundaries of Vargas. Recorded events occurred in February 1798, August 1912, January 1914, November 1938, May 1944, November 1944, August 1948, and February 1951. In the February 1798 event, flash floods and debris flows severely damaged 219 homes. Spanish soldiers barricaded an upstream-facing entrance to a fort with cannons in order to prevent debris from filling it.[2]

Prior to the 1999 disaster, the most recent major flood had occurred in 1951, but that event did not cause as much damage.[2] Based on aerial photos and records of measurements, geologists were able to directly compare the 1951 event to the 1999 event. The 1951 event involved less rainfall than the 1999 event, fewer landslides were triggered, and less fresh debris was observed on the fans.[2] The unusually strong storm in December 1999 dumped 911 millimetres (35.9 in) of rain over just a few days, triggering widespread soil instability and flow of debris.[3] Adding to the devastation, Vargas State had experienced high population growth and development since the 1951 disaster, thus increasing the toll of casualties.

Population density

Isohyet (contour of equal precipitation) map of the 14–16 December 1999 storm draped over a shaded relief map of north-central Venezuela

The alluvial fans built as sediments from floods and debris flows exit their channels and meet the oceans provide the only extensive flat surfaces along the mountainous coastline of north-central Venezuela. As such, many of them have been extensively developed and urbanized. This high population density increases the risk to life and property from flash flood and debris flow events.

As of 1999, several hundred thousand people lived in this narrow coastal strip in Vargas State. Many of these people lived atop alluvial fans formed by debris flows sourced out of the 2,000-meter (6,600 ft) peaks to their south.[4]


December 1999 was unusually wet along the north-central Venezuelan coast. The first, and less powerful, storm that month occurred 2–3 December and dropped 200 millimetres (7.9 in) of rain on the coast.[4]

Two weeks later, in a 52-hour span during 14, 15 and 16 December 1999, 91.1 centimeters (35.9 in) of rain (approximately one year's average total rainfall for the region) was measured on the north-central coast of Venezuela at Simón Bolívar International Airport in Maiquetia, Venezuela. These heavy rains included 7.2 centimeters (2.8 in) of accumulation in just one hour, between 6 and 7 AM on the 16th; precipitation on both the 15th and 16th exceeded the 1,000-year probability rainfall event. Even so, the coast received much less rain than some regions upstream.[2]

This sudden and intense storm was especially unusual because it occurred in December, while the typical rainy season in coastal Venezuela lasts from May to October. These out-of-season rains formed when a cold front interacted with a moist southwesterly flow in the Pacific Ocean. This interaction produced moderate to heavy rainfall starting in the first week of December and culminating in the 14–16 December event that caused the deadly floods and debris flows.[2]

The heaviest rains were centered around the mid-upper part of the San Julián basin, which feeds water and sediment onto the Caraballeda fan. Heavy rains persisted within 8 kilometers (5.0 mi) of the coast, and subsided on the Caracas side of the Cerro El Ávila. Rainfall rates also decreased westward toward Maiquetía.[2]



The bedrock in the region surrounding Caracas is mainly metamorphic. From the coast and extending approximately 1 kilometer (0.62 mi) inland, deeply foliated schist of the Mesozoic Tacagua Formation is exposed. Soils forming on them are fine-grained (clayey), thin (0.5–3.0 meters (1 ft 8 in–9 ft 10 in)), and often colluvial. Although the A horizon of the soil is often less than 30 centimeters (12 in) thick, the bedrock is often weathered down to greater than 2 meters (6 ft 7 in). Further inland, gneisses of the Paleozoic San Julián Formation and Precambrian Peña de Mora Formation extend to the crest of the Sierra de Avila. These units have thin soils over less-weathered bedrock; this is believed to be because of rapid erosion due to the steep slopes in this area.[2]

Because foliation planes are planes of weakness, these fabrics within the rocks strongly influence landslide and debris flow hazards.[2] Where the foliation planes are dipping towards a free surface, failure is likely to occur along these planes.[2]

Alluvial fan sedimentology and past floods

The alluvial fans that spread out into the sea from valley mouths were built by previous flood and debris flow events.[2] The modern channel systems of these alluvial fan deltas are incised into previously deposited debris flow and flood material.[2][4] Scientists at the US Geological Survey measured these old deposits. They found that they are thicker than the December 1999 ones and contain larger boulders. This means that previous debris flows were even larger than those in December 1999 and reached higher velocities.[2]

On the Caraballeda fan, the extent of the 1951 event paled in comparison to the 1999 event. Much of the deposits that constitute the Caraballeda fan are of a thickness similar to those produced in the 1999 event and contain boulders of a size similar to those observed in 1999.[2]

Oversteepened hillslopes failed during the rainstorm, sending landslides of soil into the channels (such as the braided river at bottom) and supplying sediment to flash floods and debris flows. The transmission tower on the right side of the image is 30 meters (98 ft) tall.

The USGS geologists found paleosols with organic material above and below a 10-meter (33 ft) thick layer of debris flow deposits. The bottom paleosol was radiocarbon dated to 4267 ±38 years Before Present (BP), and the top one was dated to 3720±50 years BP. This means that, at least in this area, the bed aggraded 10 meters (33 ft) in 550 years, for an average rate of about 1.8 cm (0.71 in) per year (though the aggradation occurs only during short-lived events). The scientists were not able to tell whether the deposits were from a single debris flow or multiple events.[2]

Surficial geology and geomorphology

Alluvial fan deltas in this region have shallow slopes. They are poorly channelized because sediment is being added to them from upstream (infilling the channels) at a rate equal to or greater than the rate at which it can be removed.[2]

Hillslopes are steepened past the angle of repose for noncohesive materials. This oversteepening is more than could be provided for by the frictional resistance of the sandy soils. Internal soil cohesion, negative pore pressure ("soil suction"), soil structure, and/or tree root reinforcement may be responsible for this.[2]


Terraces containing previous debris flow deposits are now perched 10–20 meters (33–66 ft) above the modern stream channels. Erosion from the 1999 flood exposed bedrock benches 50 centimeters (20 in) to 2 meters (6 ft 7 in) above the present channel. These abandoned high surfaces suggest recent and continuing tectonic uplift of the Venezuelan coast and corresponding river channel incision. In spite of the fact that most onshore faults active in this region during the Quaternary are mapped as right-lateral strike-slip, it is possible that there is a vertical component of offset in offshore faults.[2]

A 2.9-meter (9 ft 6 in) thick debris flow deposit from December 1999 is exposed by river incision during late-stage floods.