yould use indirect observation with the THERESA system. It was develop by Inra for estimating the water content of vertisols in the field by measuring the vertical deformation of the solid: http://www.sciencedirect.com/science/article/pii/037837749501165F.

Cheers, Denis

Dear Johan,

Thanks a lot for asking this interesting question. As Denis mentioned the best method is using indirect methods for evaluating crack dynamics. I suggest to obtain a relation between soil water content for a particular soil type of an area and cracking behavior of the soil.

Thanks for the answers so far!

Denis - I will check the papers that you have suggested. On a first glance, it was not so clear to me how monitoring vertical deformation can help to determine crack dimension. COuld you expand on your ideas?

Amir - This is indeed what needs to be done. However, to determine this relationship, crack properties and soil water content need to measured. I worked a lot on soil water content measurements, but I would like to have feedback on how to best determine soil crack properties, such as width, volume, and depth.

Two articles for you

http://onlinelibrary.wiley.com/doi/10.1002/hyp.5609/abstract

http://journals.lww.com/soilsci/Abstract/1993/12000/Dimension_and_Volume_of_Cracks_in_A_Vertisol_Under.7.aspx

Dear Johan,

To use vertical measurements to estimate crack volume, most people use the geometric shrinkage factor and assume that rs = 3 (isotropic shrinkage) (like the Arnold paper listed above). This assumption is subject to some debate. Here's a good discussion of this approach: http://www.hydrol-earth-syst-sci.net/17/1933/2013/hess-17-1933-2013.html

The other paper mentions filling the cracks with sand. Plaster of paris, liquid latex and cement have also been used. However, these methods are all destructive, and do not really allow for monitoring for temporal dynamics. We did some preliminary work on using a viscous solution to estimate crack volumes, but (admittedly) still need to do more field-scale validation:  https://www.soils.org/publications/sssaj/abstracts/78/2/445

Good luck!

- Ryan

Dear Mark and Ryan,

Thanks for the suggestions! I am well aware of the destructive methods, and I consider this to be the most reliable measurement, although unfortunately only for a single measurement (except when a viscous liquid is used - nice). I was not fully aware of using vertical deformation to estimate crack volume. Intiutively, I would have expected that this approach is not very accurate, except perhaps to determine the area of cracks at the surface. I will need to study the references above in more detail.

@Ryan - I loved the crack-o-meter stuff. We have been in contact about this by E-mail through Lutz Breuer. Thanks for your support.

Best,

Sander

Dear All: You seem to be hooked on the macropore myth. What you really want to know: When does a previously submerged rice-paddy soil permit aeration to what depth in the soil profile?  Cracks, fissures, voids, burrows, root channels etc. are the visible structures thought to be in charge. Why only them? Let’s turn functional: Rapid infiltration and, in particular, rapid drainage should be reasonable measures for aeration.  Rapid means  0.1  to  10 mm/s, that is  0.36 to 36 m/h. The experiment is simple: Rain on such a soil and measure water content increases as function of time at various depths in the profile, for instance, with TDR-equipment. Program your logger to record all data as fast as, or even faster than, once per minute.  Wetting front velocity is uniquely related to film thickness when assuming Stokes flow according to GERMANN and AL HAGREY (2008). Water films are in the thickness range of 10 to 100 micro meters (macropores adieu). The amplitude of the water content wave is related to the porosity that permits rapid flow. Special attention is required in vertic soils because of their swelling, thus short water pulses are required.

There will be soon a book telling you all you need:

P.F. Germann (2014) PREFERENTIAL FLOW – STOKES APPRAOCH TO INFILTRATION AND DRAINAGE. Geographica Bernensia, Bern (Switzerland). (See also: http://www.geography.unibe.ch)

_______________________________________________

Germann, P.F., and S.A. al Hagrey. 2008. Gravity-driven and viscosity dominated infiltration in a full-scale sand model. Vadose Zone J. 7: 1160-1169. Doi:10.2136/vzj2007.0172

Dear Peter,

At the risk of drifting beyond the direct topic of this question, I would be interested in hearing what exactly you mean with the 'macropore myth'. From your response, I gather that you believe that cracks do not have a major effect on the water balance of rice production systems, and  that instead you believe that fast transport in the 'matrix' should be considered. Is this correct?

By the way, I am mainly interested in infiltration losses that occur when fields are prepared for wet rice after a dry season. Besides cracks, we will also monitor water content in the soil matrix. I will pay attention to the temporal resolution.

Best,

Sander

Dear Sander:

Cracks etc. are important features that carry preferential flow, however, by far not the only ones.  Macropore Myth: Only pores that do not exert capillarity may conduct water preferentially (preferential flow, by-pass flow, non-equilibrium flow etc.). Richards’ (1931) capillary flow is sequential: Larger pores have to drain before smaller ones do, smaller pores fill before larger ones do. Preferential flow: Any pore, crack, fissure may conduct water as long as it continues over a considerable depth range. Basic assumption: Balance between gravity-drive and viscosity-control leads to constant wetting front velocities, i.e. viscous flow in permeable media (VF) based on Stokes flow. VF occurs also when capillary forces are active.

Rice paddy problem:  Take Fig. 3 in Germann and alHagrey. The recession limb expresses VF, while abstraction is matrix sorption from VF. I guess, the longer infiltration lasts the more important becomes abstraction because fissures close. Thus, I recommend to run series of short infiltration pulses.

Best wishes,

Peter