Hi! The estimate based on the grain size distribution works well for gravels and sands, a lot of empirical mathematical formuli available - just take care of the marginal conditions when selecting the appropriate formula!! Advantage is clear - money, time... but still just an estimate. However, "in situ" tests are necessary for fine-grained soils, read https://www.researchgate.net/publication/27189450_Problems_of_Hydraulic_Conductivity_Estimation_in_Clayey_Karst_Soils?ev=prf_pub.

Article Problems of Hydraulic Conductivity Estimation in Clayey Karst Soils

Brilliant thanks a million.

We are developing centrifuge methods at the moment for Irish soils. Initial results can produce a water retention curve in one day fri 0 to 15 bar. This would real solve the time issue.

As R. Adamcova has pointed out that traditional textural classification (in India we have both vernacular names and English terms for soils have different composition of , fine and coarse sand, clay, silt and gravels, boulder covered with smaller gravels, sands and clay). Maps of the soils in India are being used intensively in hydrological studies as they may easily be recombined into hydrologic soils. To collect samples from field then determine their porosity, hydraulic conductivity in the engineering lab and then genralise them using remote sensing and associations of these soils to mark their boundaries is much costly effort as uncertainty cannot be ruled out even in the core determined of zones and one governments cannot survey every parcel of land in their territory, generalisation based on climatic, vegetative and other processes have always been used to generalise soil zones of different types. Now consider again your post keeping all these points in mind, you will find you answer.

Mr Khan, I agree, everything depends on the scale of the problem and on the necessity of exact data (and money). If it is a regional problem or a local one, practical or scientific one. Within mapping bigger areas, you must generalize, so certain error is already implemented, allowing the estimate. But in silts (or even clays), one must still keep in mind, that grain size is not the only factor influencing the hydraulic conductivity - impact of the mineral composition and bulk density is extreme! So one should always make several calculations with different input data from a realistic interval and discuss the results, not just take one number from the grain size for the hydrologic study. If there is already local experience with similar soils, an analogy to data obtained from earlier field tests would be much better than the estimate from the soil structure. Of course, geodata banks are not available in all countries...If exact local calculations are needed and one has time to do it, I recommend at least some cheap field tests, e. g. by the Guelph permeameter, and extrapolation of results to relevant areas..

Mr. Fenton, I will appreciate more information on the method when it is ready to present :).

Another advantage of obtaining soil-water retention curves, is that they can be used to estimate other soil physical (mechanical) properties as well, e.g. shear strength etc.

Your question is vague and it depends on the context. What do you need these soil hydraulic parameters for ? In some cases, 'local variations' of these parameters may have dramatic consequences. A simple example is the case of levees : Very long and rather old structures, made of various soil materialsn, that may locally fail due to internal erosion, resulting in flooding. In such a case, getting real and relevant data is far too expensive ! Basically, i would thus answer that I depends on the 'risk' associated to a too simplistic estimation of the soil hydraulic parameters. And in case of doubt, consider ' safety coefficients'.

At present I am looking specifically at vertical unsaturated travel times in soil. I want to examine time lags between implementation of mitigation measures on a farm with responses in water quality.

Hi, Just two cautionary suggestion. 1. Put attention on the process you are studying. Are there preferential flow? In this case the error can be in the process schematization more than the hydraulic parameters quality. 2. Remember that the PTF are mainly derived from lab measurements (both for texture and hydraulic properties) while you want to study a field scale process. PTF smooth the real hydraulic behavior, sometimes underestimating it. Field water retention and hydraulic conductivity functions, due to hysteresis, are lower than those measured in the lab. Therefore, this two "mistakes" are sometimes compensated in the PTF, that seems to work very well. On the contrary, if the smoothing effect of the PTF is toward an overestimation of the real field behavior, the field application with the PTF will be absolutely wrong.

Mr. Fenton: be cautios with modeled hidraulic characteristics. Models as Spaw Hydrology works well in dry soils, but wet soils hidraulic properties depends greatly of soil structure and preferential flow. Here is the model

http://hydrolab.arsusda.gov/soilwater/Index.htm

Id also like to point out the issues with preferential flow.

preferential flows are not the exception that are the norm, they occur in all soil types, at all moisture contents, they are common and widespread. Have a look at

Hardie, M., W.E. Cotching, R. Doyle, G. Holz, S. Lisson, and K. Mattern. 2011. Effect of antecedent soil moisture on preferential flow in a texture contrast soil. J. Hydrol. 398:191-201.

Also any pedotransfer function based approach suffers from a high degree of inaccuracy when applied to soil physical properties that relate to soil structure, this includes, infiltration, hydraulic conductivity, field capacity, and the plant available water content. Real samples and real analysis is required for these properties.

Id also like to point out the way in which these properties can also have massive effect on the reported results, and that just because something has been measured doesnt mean it is real in a field soil. Have a look at my paper on the effects of methodology, and initial moisture content on measurement of the soil water retention curve and saturated hydraulic conductivity.

Hardie, M., S. Lisson, R. Doyle, and W.E. Cotching. 2013. Evaluation of rapid approaches for determining the soil water retention function and saturated hydraulic conductivity in a hydrologically complex soil. Soil and Tillage Research 130:99-108.

Both papers are available through researchgate.

Owen, for the problem you describe, I'd suggest measuring hydraulic properties is overkill. You can easily set a lower bound on the velocity by assuming the flux beneath the root zone is equal to infiltration minus any ET from the root zone. Beneath the root zone, the water content will adjust such that the unsaturated hydraulic conductivity is equal the flux since the gradient is unity at steady state. This flux divided by said water content is the velocity. This very simplified view of the world assumes no pref flow (which will almost surely occur) and also homogeneous conditions (which never occur), and steady state conditions. But it provides a simple lower bound to the answer. Further, by playing with the results, you will see that the velocity is not massively affected by the hydraulic properties of the soils. A coarse soil will just be drier than a comparable fine soil, but at a lower moisture content, such that the resulting velocity through the coarse soil will be higher, but not order of magnitudes higher. The bigger issue is probably trying to discern exactly what kinds of layering are present and therefore the probability of massive pref flow.

I believe there is nothing better than actually measure a particular parameter, since, to estimate a parameter, we estimate a value very close or sometimes discordant the actual value. However, the alternative estimate of a parameter may be an interesting tool, reducing the cost and time required to make the determination of a specific parameter. Based on this, I believe that after the defined objectives of the study both strategies can be empregradas. For this it is important to have clear what the precision and accuracy of the measures which the possible implications of the choice of a methodology in the final results to be obtained.

From my point of view, there any many variables to participate into the soil hydrodynamic behavior. Therefore, you need a vey low uncertainty for those hydraulic parameters that you use. Thus, a strong accuracy and reliable data can only achieve collecting field or lab real data, but even these data sometimes can not explain the soil hydraulic processes. In addition, this fact depends on the study scale and the technique or method used to measure the data in lab or field. I think a good point to start the observation could be a large hydrualic data set and comparing with the estimated and/or predicted values, especially for an specific study area.

With respect to pedotransfer functions, to decrease the uncertainty related to a single pedotransfer function, the approach of multi-model modelling as described in the papers of Guber et al. could be a possibility to deal to a given extent with uncertainties and variability.

[1] a. K. Guber, Y. a. Pachepsky, M.T. van Genuchten, W.J. Rawls, J. Simunek, D. Jacques, et al., Field-Scale Water Flow Simulations Using Ensembles of Pedotransfer Functions for Soil Water Retention, Vadose Zone Journal. 5 (2006) 234–247.

[1] a. K. Guber, Y. a. Pachepsky, M.T. van Genuchten, J. Simunek, D. Jacques, A. Nemes, et al., Multimodel Simulation of Water Flow in a Field Soil Using Pedotransfer Functions, Vadose Zone Journal. 8 (2009) 1–10.

ADV : Low time analyze , DISA : Low accuracy