GlacierTable SPHYv2.2.1

[Paul Sandoval]

Dear Wilco,

The glaciers characteristics in GlacierTable are from the first run year? I know that the logic response is yes, but are some option to give the thickness from a specific time step (I have the field information for my study area in a specific year)? Or the only way is calibrate the DDFs to adjust the thickness to this data.

Thank you!!

Cheers from Patagonia.

Paul 

[Paul Sandoval]

Dear Wilco,

Can you show me the structure of the lapse_rate.tbl table please.

Thank you.

[Paul Sandoval]

The previous question was for the next error.

[WilcoTerink]

Hi,

Sorry for my late response. Been traveling quite a bit. Yes, the ice depth in the glactable is basically your initial condition for the glacier ice depth in your model. It will then update this during each model run, depending on the melt and accumulation throughout the year.

Cheers

Wilco

[Sai Krishna]

I am looking for the process for the creation of a glacier csv table in SPHY v2.2.1 given glacier outline data (RGI v6 or v7), glacier ice thickness rasters for each glacier (say Farinotti et al.,) and glacier debris data (say Schlerer et al.,). 

I have gone through the github by Paul,(https://github.com/SQPaul/Juncal_Norte_SPHYv2.2.1/blob/main/Glaciers.ipynb), which is very useful and I thank Paul for the same. I am looking for further explanation of above code that is written especially in rationale behind creating the data that is created and some preprocessing steps that are apparently skipped (correct me if I am wrong).

Following are my questions:

1. What decides the model grid size and glacier grid sizes? Is there any criteria available in order to decide the grid sizes (or)? Does it depend on the availability of various model inputs (say soil map, DEM, air temperature, precipitation etc.)?

2. How to decide the grid size of the glacier grids? Is it based on glacier icethickness rasters (say 25m) (or) Is it based on model grid resolution (say 250m)?

3. Finally, is there a step-by-step detailed explanation for the creation of a glacier csv table, given the above requirements are set?

[Sai Krishna]

After a lot of trial and error, I could make a code that creates glacier csv file. 

In the table generated as per the SPHY v2.2.1 manual, 

1. Each row in the csv table represents a high resolution glacier grid (say 25mx25m), and given with unique ID (UID)

2. GLAC_ID column for Glacier ID string (or part of the string)

3. MOD_ID column for Model ID representing coarse resolution (say 250mx250m) model grid cell, inside which multiple glacier grids (can be upto 100) are present

4.  MOD_H column represent the coarse resolution (250m) DEM elevation corresponding to each ModelID grid cell

5.  GLAC_H column represents the fine resolution (25m) DEM elevation corresponding to each glacierID grid cell

6. DEBRIS column represents the flag for presence or absence of debris cover on a given glacierID grid cell

7. FRAC_GLAC column is by definition fraction of model ID grid cell that is covered by debris covered glacierID grid cells and debris-free glacierID cells (Need clarification in this, as explained in below query)

8. ICE-DEPTH column represents thickness of glacier in each glacier grid cell ( Need clarification in this, as explained in below query  )

Query: I have come across two type of glacier csv tables, one in the SPHY v2.2.1 manual (https://sphy.readthedocs.io/en/latest/building.html) and another in the github page of Paul Sandoval (https://github.com/SQPaul/Juncal_Norte_SPHYv2.2.1/blob/main/Glaciers.ipynb).

 I need clarification on how to assign value for fraction of glacier which represents a group of glacier cells within a Model ID grid cell, while corresponding ice-depth is going to be unique for each of the glacierID grid cell.

I see two possible cases as stated below: 

ii)Case1: Should ice-depth be averaged over each group of debris-free glacierID grid cells and group of debris covered glacierID grid cells within a given modelID grid cell? In this case, there can be variable glac_frac value for each glacierID  grid cell. 

i)Case2: Can I have unique ice-depth for each glacierID grid cell? In this case, glac_frac needs to be constant. 

I am attaching a portion of glacier csv table generated for my study area generated using Case2 process, considering fraction of glacier as constant i.e., 0.01 (which is ratio of glacierID grid area & ModelID grid cell area), to ensure I can give unique icedepth for each glacierID grid cell.

Please let me know with some explanation, on which type of glacier csv table (Case1 or Case2) is as per SPHY v2.2.1 model requirements? 

[WilcoTerink]

Hi,

I'd recommend using my GlabTop2-py python package for creating the .csv file: https://glabtop2-py.readthedocs.io/en/latest/

The choice for model grid cell-size depends on data availability, required level of detail, and computational limitations. So if your area to be modelled is huge, choosing a very small model grid cell size would result in many grid cells and making your model slow. The glacier sub-grid cell-size should be smaller than the model grid cell-size. As a rule of thumb I'd use a factor 10 smaller. So yes, if your model resolution is 250 m, then use 25 m for your glacier grids as input to the GlabTop2-py package.

Cheers

Wilco

[Sai Krishna]

Dear Wilco,

Thank you for the reply.

I have gone through the Glabtop2-py python package output csv file column names (https://glabtop2-py.readthedocs.io/en/latest/concepts.html#output)

But the column names in the Glabtop2-py (namely GLACID, Area, Zmin, Zmax, Slope, Lmax, dH, dH_dem, hmin and Tau), are not matching that of the SPHYv2.2.1 glacier csv file ( https://sphy.readthedocs.io/en/latest/building.html#sphy-2-2-0-and-newer-releases), wherein there are different column names (GLAC_ID, MOD_ID, MOD_H, GLAC_H, DEBRIS, FRAC_GLAC and ICE-DEPTH). 

Having gone through various resources available, specifically I need clarification on how to assign value for column for fraction of glacier (in glacier csv table of SPHYv2.2.1), and corresponding ice-depth for each of the glacierID grid cell as described in my earlier query.

As per my knowledge, the SPHY v2.2.1 manual is yet to be released, but I request you to provide detailed steps for creation of glacier csv table similar to the detailed steps provided for older releases (https://sphy.readthedocs.io/en/latest/building.html#older-releases).

[Sai Krishna]

I have created glacier table using procedure provided in sphy ver3 model guide(https://futurewater.gitbook.io/sphy-manual/manual/sphy-manual-3.1/build-your-own-sphy-model/glacier-table). The purpose of creating UID grid or subgrids that are smaller than model grids, to capture glacier processes at higher resolution and to resolve neighbouring glaciers. I have model domain that has glaciers (as per RGIV6.0), adjacent to each other. So, even a smaller subgrid has 2 glacier IDs. Kindly let me know if this is acceptable for the SPHY model v2.2.1 or SPHY v3, both of which have glacier table.