BEST: R estimation of Beerkan estimation of soil transfer (BEST)

In gowusu/vadose: The estimation of saturated hydraulic conductivity and soil water retention Curves in the vadose zone

Description

These functions compute the various versions of BEST. BEST computes all the 3 versions of BEST:Lassabatere etal(2006), Yilmaz et al(2010) and Bagarello et al (2014). BESTlass implements Lassabatere algorithm BESTyb implements Lassabatere algorithms of Yilmaz and Bagarello et al. et al.

Usage

BEST(data = NULL, time, I, S = 0.1, n = NULL, m = "b", y = 0.75,
  b = 0.6, r = 75, pb = 1.2, tho = 0.169, thr = 0, PSD = NULL,
  h = NULL, theta = NULL, steady = 3, init = 5, ths = NULL)

## Default S3 method:
BEST(data = NULL, time, I, S = 0.1, n = NULL, m = "b",
  y = 0.75, b = 0.6, r = 75, pb = 1.2, tho = 0.169, thr = 0,
  PSD = NULL, h = NULL, theta = NULL, steady = 3, init = 5,
  ths = NULL)

BESTlass(data = NULL, time, I, S = 0.1, n = NULL, m = "b", y = 0.75,
  b = 0.6, r = 75, pb = 1.2, tho = 0.169, thr = 0, PSD = NULL,
  h = NULL, theta = NULL, steady = 3, init = 5, ths = NULL)

BESTyb(data = NULL, time, I, S = 0.1, n = NULL, m = "b", y = 0.75,
  b = 0.6, r = 75, pb = 1.2, tho = 0.169, thr = 0, PSD = NULL,
  h = NULL, theta = NULL, steady = 3, init = 5, ths = NULL)

group.BEST(data = NULL, time, I, S = 0.1, n = NULL, m = "b", y = 0.75,
  b = 0.6, r = 75, pb = 1.2, tho = 0.169, thr = 0, PSD = NULL,
  h = NULL, theta = NULL, steady = 3, init = 5, ths = NULL, group,
  plot = TRUE, layout = c(2, 2), hlog = NULL, klog = NULL,
  opar = par(mar = c(2, 2, 1.8, 2)))

cal.BEST(data = NULL, time, I, S = 0.1, n = NULL, m = "b", y = c(0.6,
  0.8, 0.05), b = c(0.1, 0.9, 0.1), r = 75, pb = c(1, 1.8, 0.1),
  tho = c(0.01, 0.8, 0.05), thr = c(0, 0.2, 0.05), PSD = NULL, h = NULL,
  theta = NULL, steady = c(2, 10, 1), init = c(5, 7, 1), ths = NULL,
  model = "all", group = NULL)

gof.BEST(x = NULL, obs = NULL, est = NULL, P = NULL)

## S3 method for class 'BEST'
coef(object, ...)

## S3 method for class 'BEST'
predict(object, h = NULL, ...)

## S3 method for class 'BEST'
plot(x, main = NULL, xlab = "Water Content", ylab = "",
  ylab2 = "", hlog = NULL, klog = NULL, col = c("blue", "red",
  "darkgreen"), units = c("s", "mm"), mfrow = c(2, 2), type = "all",
  legend = TRUE, opar = par(mar = c(2, 2, 1.5, 2)), ...)

Arguments

data	dataframe. It can contain data with column names of "time" and "I"
time	character or numeric. The name of time variable in the dataframe. If the "data" parameter contains "time", this will be ignored. The unit must be in seconds.
I	character or numeric. The name of cumulative infiltration variable in the dataframe. If the "data" parameter contains "I", this will be ignored. The unit must be in millimetres [mm].
S	Initial value of soil sorptivity [L T^(-1/2)]. The default is 0.1. This will be optimise
n	numeric. A shaping parameter for water retention curve. This can be calibrated. If the parameter "PSD" is not set NULL lass3 will be automatically used to estimate it.
m	character. The water retention curve condition. It takes either "b" for Burdine condition or "m" for Mualem condition.
y	numeric. coefficient of BEST A parameter, commonly set at 0.75[-] This can be calibrated.
b	numeric. coefficient of BEST B parameter, commonly set at 0.6[-]. This can be calibrated.
r	numeric. Radius of infiltrometer ring [mm]. The default is 75 mm
pb	numeric. Bulk density[g cm^-3]. This can be calibrated.
tho	numeric. initial volumetric soil water content. This can be calibrated.
thr	numeric. Residual volumetric soil water content. This can be calibrated.
PSD	dataframe. Particle Size Distribution. A dataframe of two columns. The first column should be labeled "D" in the range of 0.001 to 2 mm. The second column should be named as "fr" i.e.' the fraction of Diameter. The range should be 0-1.
h	list. The range of metric head desired or measured [mm].
theta	The measured soil water content[m3m-3] at the modelled pressure levels. It is used to check the fitness of the model.
steady	The number of data points that should be used for steady state. This can be observed during infiltration measurement
init	numeric. The number of data points that should be initially used for optimisation of Best equation.
ths	numeric. The saturated soil water content [m3/m3].
group	character. The name of the group variables if the data is from different areas.
plot	whether a plot should be performed.
layout	plot layout
hlog	TRUE or FALSE. Whether metric potential should be log transformed
klog	TRUE or FALSE. Whether hydraulic conductivity should be log transformed
model	A calibration model parameter. It takes "all" for {BEST} "Lassabatere" for {BESTlass} and "Yilmaz" for {BESTyb}, " Bagarello" {BESTyb}
x	a return object of the function.
obs	Observed data
est	Estimated or modelled data
P	Number of Parameters
object	Model output object
...	Any other graphical parameter
main	Title of the plot
xlab	x label of the plot
ylab	y label of the plot
ylab2	y label of the second plot (K)
col	Color of the plot
units	Units of the plot
mfrow	The graphical layout of the plots
type	The type of plot. It takes "all" or "h"or "hk" or "k"
legend	A legend of the plot

Details

The function can perform normal BEST computation with BEST.

Value

The return parameters can be assessed with coef.BEST. The predicted soil water can be assessed with predict.BEST. In all the function returns BEST parameters A, B, C, Ks, S, slope, intercept, mod_theta, hg. Because BEST combines BESTlass and BESTyb 'l', 'y', and 'b' can be used to access Lassabatere etal(2006), Yilmaz et al(2010) and Bagarello et al (2014) parameters, respectively. For the definitions of the output of PSD see lass3. For the definitions of the output of goodness of fit tests see gof

• A: constant[L^-1]

• B: constant[L^-1]

• C: constant[L^-1]

• Ks,Ksl,Ksy,Ksb: saturated hydraulic conductivity for various algorithms [LT^-1]

• S,Sl,Sy,Sb: Sorptivity for various algorithms [LT^-0.5]

• slope,slopel,slopey: slope of steady state data for variousalgorithms

• intercept,interceptl,intercepty: intercept of steady state data for variousalgorithms

• mod_theta,mod_thetal,mod_thetay,mod_thetab: The estimated soil water content at different metric potentials for variousalgorithms

• theta: The observed soil water content at different metric potentials

• K,Kl,Ky,Kb: The unsaturated hydraulic conductivity [LT^-1]

• Kr,Krl,Kry,Krb: The ratio of K and Ks

• hg,hgl,hgy,hgb: bubbling capillary pressure for various algorithms [L]

• init: the number of data points that should be initially used for optimisation of Best equation [T]

• q: infiltration rates [LT^-1]

Author(s)

George Owusu

References

• Lassabatere, L., Angulo-Jaramillo, R., Soria Ugalde, J. M., Cuenca, R., Braud, I., & Haverkamp, R. (2006). Beerkan Estimation of Soil Transfer Parameters through Infiltration Experiments-BEST. Soil Sci. Soc. Am. J., 70(2), 521-532. doi: 10.2136/sssaj2005.0026

• Yilmaz, D., Lassabatere, L., Angulo-Jaramillo, R., Deneele, D., & Legret, M. (2010). Hydrodynamic Characterization of Basic Oxygen Furnace Slag through an Adapted BEST Method. Vadose Zone J., 9(1), 107-116. doi: 10.2136/vzj2009.0039

• Aiello, R., Bagarello, V., Barbagallo, S., Consoli, S., Di Prima, S., Giordano, G., & Iovino, M. (2014). An assessment of the Beerkan method for determining the hydraulic properties of a sandy loam soil. Geoderma, 235-236(0), 300-307. doi: http://dx.doi.org/10.1016/j.geoderma.2014.07.024

• Bagarello, V., Di Prima, S., & Iovino, M. (2014). Comparing Alternative Algorithms to Analyze the Beerkan Infiltration Experiment. Soil Sci. Soc. Am. J., 78(3), 724-736. doi: 10.2136/sssaj2013.06.0231

Examples

## Not run: 
psd=read.csv(system.file("ext","sys","psd.csv",package="vadose"))
hf=read.csv(system.file("ext","sys","h.csv",package="vadose"))
data=read.csv(system.file("ext","sys","exampleBEST.csv",package="vadose"))
mod1<-BEST(data=data,time="time",I="I",h=hf$h,PSD=psd)

pred=predict(mod1)
coefficients=coef(mod1)

# with measured theta and h
mod<-BEST(data=data,time="time",I="I",h=hf$h,theta=mod1$mod_thetab,PSD=psd,y=0.7,b=0.6)
print(gof.BEST(mod))
plot(mod,hlog="yes",klog=TRUE)
par(mfrow=c(1,2))
plot(mod,type="h")
plot(mod,type="psd")
gof1=gof.BEST(mod)
print(gof1)

#calibration
cal.mod<-cal.BEST(data=data,time="time",I="I",h=hf$h,theta=mod1$mod_thetab,PSD=psd,steady=3,
init=5,thr=0,ths=NULL,tho=0.162)
#calibrate all, set the model to 'all' Lassabatere etal(2006), 
Yilmaz et al(2010) and Bagarello et al (2014)
cal.modall<-cal.BEST(data=data,time="time",I="I",h=hf$h,theta=mod1$mod_thetab,PSD=psd,steady=3,
init=5,thr=0,ths=NULL,tho=0.162,model="all")
#get the parameters
parameters=coef.BEST(cal.modall)
S=parameters$S
A=parameters$A


## End(Not run)

gowusu/vadose documentation built on May 17, 2019, 7:59 a.m.