Valiya M. Hamza has discovered a 6,000 km-long underground ‘river' in Brazil at a depth of 4 km. He started his career in 1966 as a Senior Scientific Assistant at the National Geophysical Research Institute (NGRI) in Hyderabad, and is currently Head of the Geothermal Laboratory of the National Observatory, Rio de Janeiro, Brazil.

In an email sent to R. Prasad, he explained the basis on which the existence of the ‘river' was inferred and what sustains its continuous flow.

What is the basis on which you have confirmed the existence of an underground ‘river'?

No direct proof has been presented confirming the existence of an underground ‘river.' We have used the term “River” in a more generic sense as there are different forms for transfer of humidity in the Amazon region: atmospheric (this is transient depending on meteorological conditions), ground surface (variable fluvial discharge) and subsurface (near constant vertical seepage and lateral flows).

On the other hand, we have not found any alternative mechanism, other than large scale subsurface water flows, to account for the observed types of temperature variations.

The name “Hamza River” is not official. It was coined by my students.

Your discovery is based entirely on temperature variations found in 241 inactive oil wells. How exactly do the geothermal measurements help in obtaining information concerning groundwater movement?

The principle of geothermal method for identifying water flows in geologic media is well known in the geophysical literature.

Geothermal methods are widely recognized as suitable for mapping regional scale groundwater flows with velocities less than 10{+-}{+7}m/s. There are no other experimental techniques available for direct measurements of such small velocities.

Heat flux from the interior of the Earth towards its surface leads to a steady increase in temperatures with depth. This rate of variation is known as the geothermal gradient.

Departures of temperatures from this systematic trend are observed only under special circumstances. It may occur as a result of groundwater flow in rock formations or certain types of thermal property variations of geologic materials or subsurface thermal effects of past climate changes.

In the present case, there has been no evidence for the specific types of thermal property variations and climate changes required for producing the observed temperature changes at great depths.

This leaves water flow as the only mechanism responsible for observed temperature variations in the Amazon basin.

Non-linear changes in the vertical distribution of temperatures with depth provide clues as to the direction of water movement and its velocity.

Even if heat transfer is different in subsurface water and rocks, how do you infer the movement of water? Does moving water conduct heat in a totally different manner compared with stationary water?

Water is asubstance of relatively low thermal conductivity but high thermal capacity. In other words stationary water is inefficient in conducting heat but moving water is an excellent medium for transfer of thermal energy. This is the reason why water is preferred in cooling operations of industrial furnaces and nuclear reactors.

In geologic media water movement carries with it part of the heat content of rocks. This is known as advection heat transfer, different from the more common conduction heat transfer.

Temperature distributions in underground layers where water flows occur are different from those parts without flows.

By examining such temperature variations in boreholes or wells it is possible to identify not only the areas affected by water movement but also the direction of flow.

Does the underground ‘river' follow the course of the Amazon River?

There is no one to one correspondence between trajectories of surface rivers and underground water flows. The course of a river at the surface is determined mainly by the topography of near surface layers. On the other hand, underground flow is determined by the variations in the permeability of rock formations and local hydraulic gradient.

The statement that the Amazon River and underground water flow have the same direction is reasonably satisfactory in the context of regional scale evaluations.

Do you think that the underground ‘river' discovered has a definite starting point, and that you have identified it? What was the basis for your assumption?

Geological characteristics of the Amazon region are only poorly known. We have not been able to identify any definite starting point for subsurface flows. The area extents of the regions where oil wells have been drilled are taken as rough indicators of the zones of subsurface seepage.

Does any part of the Amazon Basin have a karst topography, and do you think some kind of a hidden fault provides for easy passage of water to great depths at the starting point?

Karst topography [typical of a limestone region] is present along the northern border of the Amazon Basin. Also, occurrences of calcareous rich formations have been identified in deep oil wells.

Hence it is possible that, at least in certain regions, deep flow of water is related to permeable calcareous rock formations.

In addition, there are a large number of east–west trending fault and fracture systems in the Amazon region. These are believed to provide paths for down flow of groundwater.

Another mechanism is seepage into soil of surface waters, there being extensive sandy soil covers arising from tropical weathering in the Amazon region. Also, the basal parts of sedimentary basins have conglomerates, with permeability comparable to or even greater than that for karst regions.

Some of the largest aquifers of the world (such as the Alter do Chao aquifer) are present at shallow depths in the middle Amazon region. Waters stored in these aquifers serve as sources of continuous seepage into lower rock formations.

According to available information on lithologic sequences, down flow of water is taking place mainly through sandy rock formations and also fracture systems.

Is it right to conjecture that all the water that is flowing in the underground ‘river' had its origin at a single point? Or do you think that water is being continuously supplied all along the course of the ‘river' by groundwater flowing down?

It is incorrect to say that flow starts at any single point. Geologic media are in general highly heterogeneous and it is difficult to identify any particular locality as a starting point. The subsurface flow receives contributions all along its trajectory until it reaches the coastal area.

Have the impermeable rocks prevented the water from flowing further down and instead directed it to flow horizontally?

The sedimentary rock formations usually have much higher permeability compared to those of basement rocks (igneous and metamorphic) beneath it. Thus vertical flow of fluids approaching the basement sections would be redirected as lateral flows.

Do the sedimentary rocks through which the water flows have a dipping gradient from west to east?

Yes. The hydraulic gradient at the surface is determined by the gradual change in topography, from relatively high level Andean foothills to the sea level at the Atlantic coast. The subsurface rock formations also follow roughly the same pattern.

Is it right to presume that the flow velocity is determined by the porosity and permeability of the rocks, which is why it flows at a slow rate?

That is correct. The magnitude of hydraulic gradient required for subsurface flow is much higher than that for surface flow.

However, the ground water velocities are small (< 10-7 m/s) and small local hydraulic gradient values may be capable of driving flows.

In addition, there are other mechanisms that may act in promoting subsurface flows such as reductions in permeability due to dissolution of natural gas and diffusion in permeable media.

Do you see uniform flow rate at all depths of the ‘river'?

The available data indicate that vertical flow rate decreases with depth. But not enough information is available at the moment for a detailed estimation of the variations in flow rate.

How deep is the ‘river'?

The Amazon River has depths varying from a few tens of metres to a few hundred metres. The flow regions of the so-called “Hamza River” have thicknesses in the range of 2,000 to 3,000 metres.

What is the volume of water flowing in the underground ‘river'?

The total volume of water in the underground ‘river' is estimated to be of the order of 1,013 cubic metres. This is not the flow rate, which is estimated to be 3,900 cubic metres/sec at the discharge zone.

Any reason why the ‘river' is so wide?

The zone of groundwater flow is estimated from information available in geologic maps and results of seismic surveys. Geological faults determine the width of Amazon basins. The width of the Amazon River is highly variable, less than 1 km in the western parts, but may reach as much as 100 km at the river mouth.

What sedimentary rock do you find at 4 km depth?

According to published information it is mostly sandstone and conglomerates.

Is the rock formation at that depth uniform all along the length of the underground ‘river'?

Most probably not. But we do not have the relevant information.

Is there any reason to worry that precious groundwater flows beyond the reach of humanity in Brazil and is of no use?

Deep groundwater flow in the Amazon region is within accessible depths for drilling. No attempt has been made to extract this, for economic reasons.

RELATED NEWS

The hidden riverSeptember 3, 2011