The fault behind the catastrophic Colombian flood

By Ross Stein and David Jacobson, Temblor

Check your flood risk

Click here for Spanish version

As a result of heavy rainfall, the Mocoa River overtopped its banks, sending water, trees, and rocks through the city of Mocoa. So far, there are 254 confirmed fatalities, with many more missing. (Photo from:


Late last week, heavy rainfall in Colombia resulted in landslides and mudflows devastating the city of Mocoa. As of this afternoon, 254 people are confirmed to have died, with many more missing. The worst natural disaster to strike the country in years occurred as rivers breached their banks, carrying rocks and trees through the city of 40,000. Because it happened at night, many people were in their homes when the deluge smashed through the town, destroying houses and carrying away cars.

In the World Resources Institute’s Global Flood Analyzer, unlike the larger Rio Caqueta to the east, Mocoa is not shown as having a flood risk. Because of the impact this flood had on the country, and perhaps its unexpected nature, we took a closer look at why Mocoa is susceptible to flooding.

Why was Mocoa flooded?

Mocoa straddles the Mocoa River, which drains the Eastern Cordillera of the Colombian Andes. Here, the ranges have been uplifted by the Mocoa Thrust Fault. In successive earthquakes over perhaps a million years, the fault has repeatedly jacked up the range front, and downdropped the basin to the south (where Puerto Guzman lies I the map below). This tends to back-tilt the stream channel, blocking its outflow from the range front to the plains. Given enough time and water, the stream would erode and deepen its valley until it forms a continuous concave-up topographic riverbed profile, erasing the effect of the faulting.

This Temblor figure shows the active tectonics around the city of Mocoa. What this illustrates is how the Mocoa Thrust Fault has uplifted and likely tilting the ranges, making the city of Mocoa more susceptible to flooding.


A competition between the fault and rain

At every range front fault traversed by a stream, there is a competition between the ability of the thrust fault to uplift the range, and the ability of the stream to erode its channel. In arid, active tectonic settings, the fault usually wins (think Nevada, Utah, or Baja California); in humid areas or where the faults have very low slip rates, the stream wins.

Despite a verdant setting, near Mocoa, the fault has prevailed, with the range front uplifted along the fault, tilted downward-to-the-north. The result is a ‘necking’ of the river valley, evident from the river terraces, which get closer to the stream as one approaches the fault. There has not been enough water coursing through the stream to fully erode the valley. At the town of Mocoa, there is a local low point where the valley widens before it necks to the south. When the river floods, Mocoa’s wide, flat expanse is inundated.

A braided river channel

Examining the Mocoa River in Google Earth, it is apparent that upstream of the neck, where the topography is flatter, the river is braided. Think of turning the water in a garden hose to the max and letting the hose drop to the ground; it snakes back and forth. In slow motion, this makes for braided channels.

This Temblor figure shows the city of Mocoa next to the Mocoa River. We have annotated both the active and abandoned river channels to illustrate why the city is susceptible to flooding.


It is also evident to us that the city is built on abandoned (former) river meanders that are not much higher than the current channel. This makes the city much more susceptible to flooding. When a river reaches flood stage, it is more likely to follow paths of least resistance, which in this case was the eastern portion of the city.

So, what caused the flood?

In the near-term: Water, and lots of it, falling in the Mocoa river catchment basin. Second, the Mocoa thrust fault, that has created a locally flat area and disturbed the river channel depth. And, finally, building a city atop old river meanders, sites of the river in the not so distant past.


CBS News
New York Times

  • Gilles Brocard

    I generally enjoy your posts, but this time you should stick to seismology.

    Mocoa is simply located where the valley widens for the first time in the downstream direction, therefore allowing for the mudflow to spread, loose energy and deposit sediments.

    This has not much to do with the fault you mention, which crosses the valley farther downstream. The fact that the river has been able to carve out a wide valley in the footwall discards your assessment that “the fault has prevailed”. The mere fact that the river is continuous across the fault trace shows that it can keep up with rock uplift, and even still has enough energy to spend on widening the valley in the footwall.

    The topography of the area shows clearly that valley enlargement is related to the river coming to flow over erodible bedrock layers, within the periclinal termination of a fold located to the NE (which a beautiful structural surface making the E side of the valley).
    What you draw as ‘terraces’ are flat irons of more resistant beds sticking out the softer surrounding layers.

  • Ross Stein

    Thank you for your comments, Gilles.

    We reached our conclusion by making a topographic profile of the stream in Google Earth. One can see that the stream has a non-equilibrium profile: The stream bed flattens more than it should near Mocoa, and steepens at the fault, before it reaches the plain. Repeated fault slip is not just a block offset at the fault, but also causes the topography to tilt down toward the north on both sides of the fault. The effects of the fault are felt well upstream and downstream, disturbing the equilibrium profile over about 15 km in either direction for a fault that extends to 15 km depth. This tilting produces a local basin or low point near Mocoa, where water can pond.

    I have not visited this site on the ground, but I believe the features south of Mocoa identified with the dashed lines are abandoned river terraces, not flat irons. Flat irons are associated with steeply-dipping strata, whereas these are at a fixed hight above the stream bed typical of river terraces. The higher terraces farther from the stream bed are presumably older.

  • Dear Ross and David,
    I have few comments on the interpretation you gave of the Mocoa flood event which partly follow those of Gilles Brocard. I will refer to your Figure 2 and 3 and I base my interpretation only from preliminary morphological observations and not knowing anything about the local geology. The Mocoa River flows along what seems to be an hogback structure, i.e. along strike of W-dipping strata, alternately less and more erodible, forming the western slope of the mountains east of the city of Mocoa (Parque Nacional Natural Serrania de los Churumbelos). This was already noticed by Gilles Brocard who defined it as a “periclinal termination” of a very large fold. I absolutely agree with him. From the topography it can be seen that south of Mocoa where you mapped the river terraces there are several N-S trending ridges having gentle western slopes and steep eastern ones. These morphological features can be seen from Google Earth images and topography, but also from the terrain map of the Temblor map service (that shows contour lines and shades). These ridges are formed by the less erodible W-dipping strata. The river follows the more incised sections, probably coinciding with softer and more erodible set of strata. What you indicate as river terraces in your Figure 2 are actually the ridges forming the hogback structure. Just downstream of the city of Mocoa, the river steps eastward from one valley to another probably following some structural control. Upstream of this change of course there is a small alluvial plain and a change of river pattern forced by the morphological neck at the crossing of the W-dipping strata.

    The city of Mocoa is founded at the confluence of two small right tributaries of the larger Mocoa River and the mood flood was actually generated by these two streams not by the Mocoa River itself. There is no abandoned paleomeander of the Mocoa River within the city as you show in your Figure 3. The houses are actually built on a set of small hills separating the two tributaries and the Mocoa River valley. The tragedy happened because there are also a lot of houses built on low-lands near the confluence of the three channels. That area was actually flooded and devastated by the mood avalanche as it can be clearly seen from many pictures found on internet. As regards the cause of the mood avalanche I read on the internet that it was probably due to wild deforestation and illegal mining on the mountain slopes west of Mocoa. The two tributaries are short and very steep so probably there was a kind of flash flood not necessarily triggered by an exceptional rain. I don’t see any possible control of the tectonic activity of the frontal thrust, in the long term, on the occurrence of this flood event.

    As regards the crossing of the Mocoa River of the frontal thrust near the town of Villagarzón, it can be seen from satellite images that the river turns abruptly eastward and its valley in that section is entrenched of about 50 meters within a gentle S-sloping former alluvial plain; there are many evidences of N-S directed anastomosing paleodrainages that could represent the former Mocoa River course. The question is: why does the Mocoa River make this turn? I really don’t know. Apparently, there is no (tectonic-) topography in front of the valley outlet. It may have been caused by a capture from the larger river flowing south-eastward near Puerto Guzmán, or it may be that the Mocoa River used to flow southward during past wetter climatic condition and there was a climatic forcing for the abandonment, or the diversion may be due to the local, tectonic induced subsidence at the base of the mountain range. By sure there cannot be subsidence right to the south of the valley outflow as in your Figure 2 otherwise the river would have been attracted towards it. I conclude saying that I do believe that tectonics in the long term controls drainage systems, and we have many examples also here in Italy, either in mountain or alluvial settings, but in this case in my opinion I think that the flooding and the victims were due to the lack of planning of the city.

    P.S.: I tried to post these comments two days ago but the DISQUS system detected it as “spam”.

  • Gabriel Veloza

    Dear Ross, I really enjoyed your post. I think it makes it clear that there is a long term link between fault movement and climate. However, for this particular case I do agree with Gilles and Pierfrancesco. On top of the geomorphic features described by G & P, I´ve have the opportunity to work in industry seismic interpretation in this zone and the fault does not extend as deep as 15 km. In this region of the southern Colombian Andes, as well as farther north, the thrust front flattens out at about 4-6 km deep. More importantly in this case to me, is the lack of policies and government action towards deforestation and illegal mining. Those are recurrent problems for almost all the rural areas in Colombia, and in this particular case it has been documented for some years now (

  • Ross Stein

    (For others, you should know that Gabriel is the leader of the Latin America active fault inventory published in GSA Today that Temblor uses in its maps, for which we are grateful.) Thank you, Gabriel. If you and others with local experience do not agree with our interpretation, we are comfortable withdrawing it.

  • Gabriel Veloza

    Hello Ross, please do not withdraw your post! I think just having the opportunity to see these issues from different angles will help to better understand the underlying processes in mountain building, landscape evolution and natural hazards. Even if we disagree on the root causes for this one, I believe that spreading some of these concepts in an understandable fashion for the general public (an area where you exceed) will help make folks aware of the impact of geological processes. Furthermore, this type of scientific debate would help policy makers to get a better grasp on the issues their communities face.