Tuesday, May 19, 2015

Inside the Research: Part 1

This is about what I call it as a paper war, between A.J Parsons and I.D.L Foster with L. Mabit, K. Meusburger, E. Fulajtar and C. Alewell. Those authors are basically a master on fingerprinting tracer. The war begins when Parson and Foster published a paper in 2011 questioning about what we can learn from using 137Cs fingerprint tracer. Two years after in 2013, Mabit and team replied by a paper stated about the usefulness of 137Cs fingerprint tracer. And again, at the same year, Parson and Foster replied the paper with the assumption of science. Here, for the beginning, I will write a resume about the first paper from Parson and Foster (2011), what we can learn about soil erosion from the use of 137Cs.
Parson and Foster (2011) were questioning about many assumption used in the techniques and lead to the conclusion that 137Cs cannot be used to provide reliable information about rates of soil erosion. Here are the assumptions which Parson and Foster (2011) were questioning:
1.        The atmospheric fallout is locally, spatially uniform
a.       Parson and Foster (2011) highlighted this point by using rainfall data and stated that there is a strong relationship between 137Cs level in soil and rainfall. They stated that it would be more reasonable if assuming a bias in rainfall amount towards the rain-bearing direction, rather than to assume a spatially uniform amount. Apart from that, soil is very heterogeneous; only one single visit sampling is no more than 5% of the variability.
b.      They stated again that the variability of 137Cs inventories are depending on the spatial variability in infiltration rates of the vegetation cover, whether within a single or variety vegetation. However, the process of 137Cs transfer from vegetation to soil is neither straightforward nor a simple one-way process.
c.       Other than that, global transport of Cs is dominantly through stratosphere, in such storm local infiltration rates are likely to be exceeded causing pounding in shallow depression and runoff on slope leading to 137Cs redistribution during the transfer process in the soil through runoff. Thus associated with the movement of water carrying the 137Cs from atmosphere and vegetation.
These three main points are always being assumed by most of the scientist who are conducting this kind of research. Owens and Walling (1996) suggest taking three cores within 1 square metre and bulking the samples to obtain a better estimate of the mean reference value than a single sample. Single sample will not provide any statistically valid method to estimate the reliability of the estimate of the true reference mean value. Nor will it provide any data that can be used to assess the significance of differences from the estimate of the reference value. The accuracy for all measurements is necessary needed.
2.        The fallout is rapidly and irreversibly fixed onto spatially uniform
The next assumption is the elimination of plant uptake and both micro and macro organism role as well as soil particle itself on the adsorption of 137Cs. Following the figure below is a pathway of 137Cs movement through he soil-plant system which is somehow eliminated by the researcher.

Fig. 1. Pathway of 137Cs movement through the soil-plant system
Removal of plants on which 137Cs is initially deposited either by grazing or harvesting before washoff will reduce the local inventory, as will subsequent removal of plants into which 137Cs has been taken up.
3.        The subsequent redistribution of fallout is due to the movement of soil particles
By this assumption, it will lead all the calculations become overestimated, somehow underestimated. This assumption calculated soil redistribution rates and a sediment delivery ratio without accounting for the effects of 137Cs losses via this pathway. If significant amounts of fallout are preferentially adsorbed onto soil particles being carried by runoff during deposition events then the value from the reference sites will be an overestimate of the initial inventory from potentially eroding sites and probably an underestimate of that from potentially depositing sites.
Water is not the only available agent of soil erosion. Wind erosion and frost erosion, particularly on bare agricultural land, may play a significant role in soil and hence 137Cs removal. There are few data of comparison between wind and water erosion rates directly as well as with frost erosion.
4.        Estimates of soil erosion can be derived from measurement of 137Cs inventories
Inventories of 137Cs do not provide measures of rates of soil erosion directly, but need to be used in conjunction with some form of conversion model. There are two broad measure categories: empirical and theoretical. The value of empirical relationships lies not so much in converting measures of 137Cs to rates of soil erosion as in their ability to demonstrate that 137Cs losses can be used as a surrogate for erosion measurements. Moreover, theoretical conversion models which exist in a variety of forms and complexity as well as the many default parameters needed, typical values leads from the difficulty of obtaining site-specific values. The use of such default parameter values compounds uncertainty in estimates of erosion rates.
5.        Error margins of 137Cs inventories
There are four sources of error associated with determining the 137Cs inventory for a specific site:
a.       associated with the sampling technique
It might be argued that the scraper plate method, which samples a larger area than the cylindrical corer or box corer, would reduce the error in the estimation of dry bulk density. However, Parson and Foster (2011) stated that they have found no evidence in published studies as to the benefits and pitfalls of using these different methods or of the likely impact of sampling methods on the magnitude of inventory errors.
b.      associated with the fractionation of the sample before analysis
The most recommendations from mainstream researchers is to use fractionation of soil to < 2 mm before counting. However, some researchers are reported that the > 1.0 mm fraction of a soil may contain up to a third of the total 137Cs activity. Yet, sample processing is therefore also likely to have a significant impact on the calculated inventory.
c.       associated with measurements of the random process of radiometric decay
Most published studies using 137Cs use gamma spectrometry to determine sample activity. All detectors require energy and absolute efficiency calibration. However, several uncertainties in measurement accuracy exist, including correction for background radiation levels, packing geometry and self absorbance.
d.      Associated with determining the limits of detection.
The mainstream theory is that by increasing count times will reduces the error of the count. In fact, 137Cs was not detectable in a sample after a count time of 28,800s (using a thin disc geometry). Increasing the count to 86,400s somehow gave a counting error of 11.66% but increasing the count time to 240,000s reduced this error to ca. 6%. These errors are a direct function of the uncertainty in the radioactive decay process.
The error in measurement is therefore a function of radionuclide activity and count time yet few studies explicitly specify an acceptable error and adjust count times to achieve this.
6.        The accuracy of the understanding of soil erosion processes
a.       Oversimplification of particle-size effects on the estimation of erosion rates.
The strong argument is both deposits of eroded soil and samples taken from the outlets of runoff plots fail to take into account the effects of differential travel distances of particles of different sizes. using both theoretical arguments and empirical data, that eroded soil travels a finite, and typically short distance. Furthermore, travel distance increases as particle size diminishes. Finer sediment will leave catchments and coarser sediment will be deposited within them. If we failure to take account of this difference, it will lead to errors in sediment budgets based on 137Cs measurements.
b.      Pertains to erosion on agricultural land is the relationship of the timing of erosional events to the timing of cultivation.
Thus, depends on the relationship between 137Cs fallout, erosion and ploughing. It will make the actual value likely falls somewhere between these two extremes.

            It can be conclude that many assumptions are very risky to hold the experiment. Apart from that, this technique needs to have more development, particularly to eliminate the number of default assumptions. There is no 100% perfect researches even a Nobel Prize man. This kind of corrections and critics will lead to more other new improvements to become the better standard technique. 

What can we learn about soil erosion from the use of 137Cs?
A.J Parsons and I.D.L Foster
Earth-Science Reviews 108 (2011): 101-113