GoF in PlotBasinOutput does not correspond with GoF from HYPE output.

[MaartenWynants]

Hej,

In certain model settings, the GoF outputs from HYPE do not correspond with these I get when plotting these in PlotBasinOutput. See examples: simass.txt In earlier HYPE setting these were exactly the same, so I was wondering what could explain these differences and what GoF I should trust here. Especially the difference in NSE is large. I saw in the code that hydroGOF gof gof.default is imported, but not sure what package is actually used? Are the values calculated again or just pasted from the HYPE outputs?

Regards, Maarten

[rcapell]

Hi Maarten, thanks for the report! I haven't run any tests yet, so cannot confirm your observation yet. A quick first reply:

• Performances given in the plot are calculated from the basin output file using function hydroGOF::gof
• Performances are calculated for the plotted period, so there could be a difference if you selected a limited time period to plot
• simass.txt gives aggregated performances for the model domain, for individual gauges, subass.txt gives results for each (could be the same though if your domain contains exactly one gauge)

/rené

[MaartenWynants]

Hej Rene,

As a quick response, the plotted periods are the same so it is not that. Also the domain only contains one gauge so that should be the same. Below is the subass for TS: subass6.txt I will have a look at this HydroGOF::gof package and a further look at the code, maybe there is an issue with what data it compares.

[rcapell]

Sounds good, thanks! One last question: which HYPE version are you using?

[MaartenWynants]

HYPE_5_18_0

[rcapell]

Now I got to look at this a bit. I cannot reproduce the problem. Attaching some test data with COUT/ROUT time series and an accompanying subass file. With

library(HYPEtools)
library(hydroGOF)

# import test results
bsn <- ReadBasinOutput("testbasinoutput.txt")
sbs <- ReadSubass("testsubass.txt")

NSE(bsn$COUT, bsn$ROUT)
# => -1.591414
KGE(bsn$COUT, bsn$ROUT)
# => -0.1366507
sbs$NSE
# => -1.5914
sbs$KGE
# => -0.1367

PlotBasinOutput(bsn, driver = "screen", area = 873*10^6)

NSE(bsn$COUT[1:100], bsn$ROUT[1:100])
PlotBasinOutput(bsn, driver = "screen", area = 873*10^6, from = 1, to = 100)

Everything seems to be ok. Could you cross-check/am I missing something? HYPE 5.18.0 HYPEtools latest master hydroGOF 0.4.0

testbasinoutput.txt testsubass.txt h

[c-brendel]

Closing as non-reproducible after conversation with Maarten