Th
e Ca l
![](http://www.sci<font color=)
enc
edir
ect.com/scidirimg/
entiti
es/2bb.gif" alt="modifi
er l
ett
er turn
ed comma" bord
er=0>Isard catchm
ent (1.32 km
2), a sub-basin of th
e Vallc
ebr
e exp
erim
ental catchm
ents, yi
elds larg
e amounts of s
edim
ents (about 580 Mg km
− 2 y
ear
− 1) that ar
e produc
ed in r
elativ
ely small but v
ery activ
e erod
ed ar
eas (badlands). S
ev
eral lin
es of
evid
enc
e sugg
est that th
er
e is a d
elay b
etw
een s
edim
ent production, caus
ed by int
ens
e summ
er rainstorms, and s
edim
ent transport, occasion
ed by th
e main floods produc
ed by larg
e pr
ecipitation
ev
ents following w
et ant
ec
ed
ent conditions. First, a calibration–validation
ex
ercis
e was carri
ed out with s
edim
ent yi
eld data obtain
ed using contain
ers provid
ed with slot divisors in a badlands micro-catchm
ent (1240 m
2). Th
en, th
e mod
el was appli
ed to th
e main badlands ar
eas in th
e Ca l
![](http://www.sci<font color=)
enc
edir
ect.com/scidirimg/
entiti
es/2bb.gif" alt="modifi
er l
ett
er turn
ed comma" bord
er=0>Isard sub-catchm
ent for a 4-y
ear p
eriod and th
e simulat
ed s
edim
ent yi
elds w
er
e compar
ed with th
e r
ecords at th
e gauging station. Th
e t
est was p
erform
ed with th
e G
en
eraliz
ed Lik
elihood Unc
ertainty Estimation (GLUE) approach for ass
essing th
e unc
ertainty associat
ed with mod
el pr
edictions, which assum
es that many param
et
er s
ets can giv
e acc
eptabl
e simulations. Th
e r
esults d
emonstrat
ed th
e capacity of KINEROS2 to simulat
e badland
erosion, although it show
ed limit
ed robustn
ess. A cl
ear t
emporal mismatch b
etw
een
erosion and s
edim
ent transport and th
e r
el
evanc
e of s
edim
ent stor
es in th
e catchm
ent w
er
e confirm
ed, whil
e th
e total w
eights of s
edim
ent w
er
e g
en
erally und
er-pr
edict
ed. Th
e limit
ed suitability of th
e ar
ea us
ed for calibration or th
e rol
e of s
edim
ent sourc
es not simulat
ed in th
e approach may account for this shortcoming.