Strange values in ancient Tulare Lake

[dylanbeaudette]

Are these problems, or just the unique landscape position and thick lacustrine deposits?

• https://casoilresource.lawr.ucdavis.edu/gmap/?loc=36.03022,-119.92144,z13
• https://casoilresource.lawr.ucdavis.edu/sde/?series=tulare

[dylanbeaudette]

Re-visiting the 2021 data, and comparing with SSURGO / KSSL. Look OK to me.

• KSSL: ~ 50 kg / sq. m.
• SSURGO: ~ 59 kg / sq. m.

library(aqp)
library(soilDB)

# KSSL
x <- fetchKSSL(series = 'tulare')
# SSURGO
y <- fetchSDA("compname = 'tulare'")

# grouping ID for later
site(x)$origin <- 'KSSL'
site(y)$origin <- 'SSURGO'

# normalize select hz properties
x$db_thirdbar <- x$db_13b
y$db_thirdbar <- y$dbthirdbar_r

x$om <- x$estimated_om
y$om <- y$om_r

x$name <- x$hzn_desgn
y$name <- y$hzname

# combine
z <- aqp::combine(x, y)

# no fragments
horizons(z)$soil_fraction <- 1

# horizon SOM
hd <- horizonDepths(z)

z$SOC_kg_sq_m <- profileApply(z, FUN = function(i) {

  # horizon thickness
  i$thick <- i[[hd[2]]] - i[[hd[1]]]

  # horizon SOM kg / sq. m.
  # horizon thickness (cm) * OM fraction (0-1) * soil fraction (0-1) * Db 1/3 bar (g/cm^3) * unit conversion
  res <- i$thick * i$om / 100.0 * i$soil_fraction * i$db_thirdbar * 10

  return(res)
})

# sum
z$sum_SOC_kg_sq_m <- profileApply(z, function(i) sum(i$SOC_kg_sq_m, na.rm=TRUE))

# simple plot
par(mar=c(0,0,3,1))
groupedProfilePlot(z, groups = 'origin', color = 'om', group.name.offset = -20, print.id=TRUE)
groupedProfilePlot(z, groups = 'origin', color = 'db_thirdbar', group.name.offset = -20, print.id=TRUE)
groupedProfilePlot(z, groups = 'origin', color = 'SOC_kg_sq_m', group.name.offset = -20, print.id=TRUE)

# summarize: note that one of the SSURGO components is bogus: missing lower horizon data
s <- site(z)[, c('origin', 'sum_SOC_kg_sq_m')]
tapply(s$sum_SOC_kg_sq_m, s$origin, median)

[dylanbeaudette]

1/3 bar Db values seem way too low.

https://casoilresource.lawr.ucdavis.edu/soil-properties/?prop=db&lat=36.0602&lon=-119.7867&z=9

Checking the current SSURGO data, Tulare components seem too low.

library(aqp)
library(soilDB)
library(sf)
library(terra)
library(viridisLite)

## investigate Db at 1/3 bar

x <- fetchSDA("mukey = '461824'", duplicates = FALSE)

horizons(x)$.hzlab <- sprintf("%s\n%s g/cc", x$hzname, x$dbthirdbar_r)

.a <- list(
  label = 'compname', 
  shrink = TRUE, 
  scaling.factor = 1.2, 
  name.style = 'center-center', 
  cex.names = 0.75
)

options(.aqp.plotSPC.args = .a)

par(mar = c(2, 0, 3, 2))
plotSPC(x, color = 'dbthirdbar_r', name = '.hzlab', col.label = 'Bulk Density at 1/3 Bar (g/cc)')

.txt <- sprintf("National MUSYM: %s", unique(x$nationalmusym))
mtext(.txt, side = 1, at = 6)

[dylanbeaudette]

The 1/3 bar Db values in KSSL are surprisingly low.