Driving Noah-MP single point simulation using tower-observed data?

[tais958]

Hi, Maybe I had missed this part in the user guides you provided, but I wondered if LIS allows user to conduct single point simulation using Noah-MP and potentially makes it driven by observed met forcing? Thank you in advance for any response!

Sheng-Lun

[dmocko]

Hi @tais958

It's not possible to conduct a truly single-point simulation, but we can simulate a 2 grid point by 2 grid point domain.

For an example, please see the Bondville testcase: https://github.com/NASA-LIS/LISF/tree/master/lis/testcases/metforcing/Bondville

NOTE - we don't keep these testcases updated to the current software. Please follow the Users' Guides for constructing your config files. But you can get a sense of what is being done from the ldt.config and lis.config files in this directory.

[tais958]

Hi @dmocko

Thank you so much for the comment. I noticed that Bondville forcing needs to be recorded on local time, and I made the adjustment before running the simulation. However, as you can see in the figures, even though the forcing variables (examined wind, humidity, temperature, pressure, radiation, etc..) extracted from the LIS (S10) outputs follow what pulled from the forcing ASCII txt file, the simulated surface fluxes seem to be shifted by around +6 hours. Do you know of any other setting may affect how model acts in terms of diurnal cycle? Much appreciate it!

[dmocko]

HI @tais958

The point in my previous post is that the LIS forcing reader for Bondville automatically converts the Bondville forcing from local time to UTC time. The LIS output from this run will be in UTC time, which is why you are seeing the shift in the simulated surface fluxes output.

[tais958]

Hi @dmocko

Thanks for fast response. I am a bit confused as I've been using the same workflow to process the two LIS run outputs (either NLDAS2 or S10). I'd assume they all recognize the time coordinates in the LIS outputs are in UTC. In this case, I didn't do any conversion in time for S10. Does this make sense to you? Please advice if I misunderstood anything here. Thank you!

[dmocko]

Yes for S10 and NLDAS-2 have outputs in UTC. Maybe in your plots you are shifting the data if you think the outputs should have been in local time?

[tais958]

Hi @dmocko

I confirm no time shift has been applied to the plotting. Upon checking timeseries plots for SW and SHF, it seems this is not purely time shift issue but related to some suppression in diurnal radiation in the morning. As the figures shown below, while LIS(S10) SW has a reasonable diurnal cycle over the days inspected, SHF stays low for ~6 hours in the morning before it goes up. Then, at ~18 UTC, all three curves decline at a similar rate. Given these, I suspect there's some part of scripts may also act to cause this issue. Would you think of places might worth checking?