Modify urban coefficients from Åstebøl et al. (2012)?

[JamesSample]

The notebook here uses coefficients from Åstebøl et al. (2012) to estimate urban inputs that vary with runoff (rather than the "static", area-based inputs used in the old version of the model).

Åstebøl et al. provide two sets of coefficients: one for "tett by" and one for "åpen by", but the AR50 land cover dataset used by TEOTIL only defines one urban class, described as "Bebygd: Boligfelt, tettsted, by, samferdsel, industriområde o.l.".

As a starting point, I have used the coefficients for "tett by", but the plots in this notebook suggest urban fluxes estimated using these coefficients are much higher (5 to 10 times higher) than urban fluxes in the original model. Using the coefficients for "åpen by" instead may therefore be more appropriate, but I'll wait until I can compare the model output to measured fluxes from Elveovervåkingsprogrammet before making a final decision.

[JamesSample]

Based on further testing, I think the coefficients for "tett by" from Åstebøl et al. are much too high. I have therefore switched to using the "åpen by" coefficients in db44e31. However, preliminary testing suggests that even these may be too high.

Leaving the issue open as it needs further investigation.