In AustralianAntarcticDivision/ZooScatR: ZooScatR - Acoustic Scattering of zooplankton using DWBA

knitr::opts_chunk$set(echo = TRUE)
library(ZooScatR)
library(reshape2)
library(ggplot2)

Create a TS simulation and plot the results

Data preparation

First let's load standard parameter file, included in ZooScatR. We will then make minor adjustments to include a more realistic krill shape, simulate data from 18 to 300 kHz, reduce the number of integration elements and frequency steps,as a compromise between speed and data resolution. We also have to set the soundspeed in the surrounding medium. Finally we will run a simulation with the standard distributions for Lengths and Oirentations.

fname <- paste0(system.file(package="ZooScatR"),"/extdata/configs/config_0.dat") #Loacation of the parameters file
para = read_para(fname) #Read parameters file
para$simu$var0 = 18 #simulate from 10 
para$simu$var1 = 300 #...to 300 kHz
para$simu$ni = 200 #resudce the number of elements and frequencies to improve speed
para$simu$n = 283

#Create list with soundspeed info
misc <- list(cw=1500)

#change to the more realistic krill shape
profname <- paste0(system.file(package="ZooScatR"),"/extdata/profiles/euphaus1.dat") #krill example
para$shape$prof_name <- profname

#make a length simulation
para$shape$ave_flag = 1

#make an orientation simulation
para$orient$ave_flag = 1

Simulation

Now we are ready to run the simulation. We fix the number of simulations with nang for orientation angles and nl for lengths.
Note that we set simOut to TRUE (which strictly speaking is not necessary as this is the default setting now):

#Run DWBA based on config file
res <- bscat(para=para, misc=misc, simOut = TRUE, nang=100, nl=100) #Target strength vs Frequency

The way the simulations work is that first simulations for the orientation distribution are generated. These are then stored as an output and the mean is computed for furhter simulation with the length distribution. Alternatively, individual simulations could be run for a set of lengths and orientations.

Once the simulations are completed we can have a look at the output.

• res$y is the standard averaged output result

• res$var contains the Frequencies for which the simulations were run

• res$ysim contains the simulations for the diffrenet orientations

• res$ang contains the orientation angles for which the simulations were run

• res$ysimL contains the simulations for the different Lengths, based on the average orientation result.

• res$Length contains the lengths for which the simulations were run

Plots

This allows us to generate easily compute some overview plots, that might be informative.

Orientation simulations

o_sim = melt(res$ysim)
o_sim$theta = res$ang[o_sim$Var2]
o_sim$Frequency = res$var[o_sim$Var1]
o_sim=o_sim[,3:ncol(o_sim)]
names(o_sim)[1] <- 'TS'

mTS = data.frame(Frequency=res$var,TS=res$y)

ggplot()+geom_line(data=o_sim, aes(x = Frequency, y=TS, group=theta),lty=2, lwd=0.5, alpha=0.4)+
  geom_line(data=mTS, aes(x=Frequency, y=TS), lty=1, lwd=2)+theme_classic()+theme(text=element_text(size=14))

ggplot(data=o_sim, aes(x=Frequency, y=theta, fill=TS))+
  geom_raster()+
  scale_fill_viridis_c()+
  scale_y_continuous(expand=c(0,0))+
  scale_x_continuous(expand=c(0,0))+
  theme_classic()+theme(text=element_text(size=14))

Length simulations

l_sim =  melt(res$ysimL)
l_sim$Length = res$L[l_sim$Var2]
l_sim$Frequency = res$var[l_sim$Var1]
l_sim=l_sim[,3:ncol(l_sim)]
names(l_sim)[1] <- 'TS'

ggplot()+geom_line(data=l_sim, aes(x = Frequency, y=TS, group=Length),lty=2, lwd=0.5, alpha=0.4)+
  geom_line(data=mTS, aes(x=Frequency, y=TS), lty=1, lwd=2)+theme_classic()+theme(text=element_text(size=14))



ggplot(data=l_sim, aes(x=Frequency, y=Length, fill=TS))+
  geom_raster()+
  scale_fill_viridis_c()+
  scale_y_continuous(expand=c(0,0))+
  scale_x_continuous(expand=c(0,0))+
  theme_classic()+theme(text=element_text(size=14))

AustralianAntarcticDivision/ZooScatR documentation built on Aug. 13, 2022, 1:21 a.m.