Sub-daily Run

[Nooshdokht-Bayatafshary]

Dear developers,

I am currently working on incorporating hourly meteorological data, including Precipitation, Temperature, and the necessary information for calculating ET0 from LISVAP, into my LISFLOOD model. However, I would appreciate clarification on a few aspects outlined in the model manual.

As per my understanding from the manual, the LISFLOOD model typically processes daily data. Could you please clarify how the LISFLOOD model processes daily data and if there is a mechanism for converting it into hourly data? I am particularly interested in understanding how the model handles the transition from daily to hourly data and whether it supports the input of hourly meteorological data.

Furthermore, I am curious to know if the LISFLOOD model can handle and utilize hourly meteorological data directly. If this is feasible, should the hourly data be entered into the LISVAP module for ET0 calculation, as well as for Precipitation and Temperature inputs for the LISFLOOD model? My concern arises from the information in the manual suggesting that using hourly data may introduce errors in snow melting calculations. Could you guide how to appropriately run and calibrate the model with hourly ground data to mitigate these errors?

Lastly, I would like to inquire about the units of meteorological data when running the model hourly. Clarification on this matter would greatly assist me in ensuring accurate data input.

Thank you for your time and assistance. I appreciate your support in addressing these queries.

Best regards, Nooshdokht

[StefaniaGrimaldi]

Dear @Nooshdokht-Bayatafshary ,

We are pleased to know that you are using OS LISFLOOD and we are keen to collaborate with you. Please find below a brief answer to all your questions.

OS LISFLOOD can be readily used with sub-daily input meteorological data, and it can be calibrated using sub-daily discharge data. In fact, the European Flood Awareness System of the Copernicus Emergency Management Service (CEMS EFAS, https://confluence.ecmwf.int/display/CEMS/Latest+operational+EFAS+release) makes use of OS LISFLOOD with 6hourly time steps. In the CEMS EFAS implementation, DtSec is 21600 and DtSecChannel is 3600.

The input meteorological data for sub-daily implementations must be referred to the desired sub-daily frequency. However, care must be taken to use the appropriate units (the set up might not be intuitive!). Total precipitation: you can use hourly steps (24 steps in a day) but the units must be mm/day. The model internally makes the adequate conversion. Reference evaporation and evapotranspiration: you can use hourly steps (24 steps in a day) but the units must be mm/day. The model internally makes the adequate conversion. Average temperature: you can use hourly steps (24 steps in a day), in this case, each value is the hourly average value.

LISVAP is expected to run with hourly steps. However, minor adjustments could be needed in the writing routine. Please let us know whether you intend to pursue this path and we will do our best to prioritize the checks.

Snow melt computations are based on degree-day methodology, therefore, we recommend to evaluate the accuracy for specific application with sub-daily steps.

We hope that we were able to clarify some of the doubts, and that our answers allow to make progress in our conversation. Kind regards, Stefania (on behalf of the OS LISFLOOD developers team)

[Nooshdokht-Bayatafshary]

Dear Stefania and the OS LISFLOOD Developers Team,

Thank you for your prompt and detailed response. Your insights have been invaluable, and I appreciate your willingness to collaborate.

I am indeed interested in utilizing LISVAP at hourly steps and would like to inquire about the recommended setting for DtSec. Should I set DtSec as 3600, aligning with a one-hour time step for the LISVAP module? Alternatively, would it be more appropriate to set DtSec as 86400 (with daily meteorological inputs) and configure ReportSteps for hourly outputs, specifying the number of hours I desire maps for (e.g., 1..[number of hours])? Based on my understanding of the LSVAP's manual, some of the simplifying assumptions made in LISVAP related to the radiation balance are not valid at time steps smaller than days. Therefore, it is advised to use LISVAP for daily time intervals only (i.e. DtSec should always be 86400)

Additionally, as I plan to proceed with sub-daily implementation, could you kindly guide me if any other adjustments or modifications are necessary in the setup? Your expertise in this matter will greatly assist me in setting up my model with sub-daily meteorological data. I am keen on pursuing this path and would appreciate your continued support in prioritizing the necessary checks for LISVAP running at hourly steps.

Once again, thank you for your assistance.

Best regards, Nooshdokht

[StefaniaGrimaldi]

Dear @Nooshdokht-Bayatafshary,

thank you for confirming your plan to use LISVAP with hourly time steps.

In order to run LISVAP with hourly time steps, it is necessary to set DtSec=3600 and to prepare all the input meteorological data (e.g. temperature, wind speed, solar radiation, etc..) with hourly time steps. As mentioned in the previous message, we will make the required checks to make sure that LISVAP correctly reads and writes the hourly steps. We recommend following up on this task within the LISVAP GitHub repository (https://github.com/ec-jrc/lisflood-lisvap).

Concerning the preparation of the meteorological forcings, you might be interested in the pyg2p interpolation and post-processing tool: https://github.com/ec-jrc/pyg2p. This tool allows to interpolate in-situ meteorological observations or weather forecasts (GRIB files) to the desired grid resolution and coverage. The tool also allows to implement post-processing steps such as temperature correction using elevation.

LISVAP allows to use different sets of meteorological variables. The set of variables of your choice can help assess the reliability and impact of the simplifying assumptions. For instance, the computation of daily extra-terrestrial Angot radiation to assess net longwave radiation.

We hope that this answer helps, we will surely let you know when the checks on LISVAP have been completed.

Kind regards, Stefania

[StefaniaGrimaldi]

Dear @Nooshdokht-Bayatafshary,

we suggest closing this issue and continuing the discussion within the LISVAP GitHub repository.

Please do not hesitate to let use know whether you have questions regarding LISFLOOD,

Kind regards, On behalf of the developers team, Stefania