Skip to content

Latest commit

 

History

History
102 lines (77 loc) · 4.34 KB

File metadata and controls

102 lines (77 loc) · 4.34 KB

QuickView has been developed using EAM's history output on the physics grids (pg2 grids) written by EAMv2, v3, and an intermediate version towards v4 (EAMxx). Those sample output files can be found on Zenodo.

Developers and users of EAM often use tools like NCO and CDO or write their own scripts to calculate time averages and/or select a subset of variables from the original model output. For those use cases, we clarify below the features of the data format that QuickView expects in order to properly read and visualize the simulation data.

!!! tip "Tip: Consistency Between Simulation File and Connecitivity File"

One of the repeatedly encountered causes of error when loading files
in QuickView
is that the grid described by the connecitivity file does not match
the grid in the simulation data file.
If the `Load Files` button in the [toolbar](toolbar.md) has been clicked
but the status icon continues to show a red circle with an exclamation mark
after many seconds of time, the first thing we suggest
checking is whether the two specified files correspond to the same grid.

!!! warning "Caution: Newer EAMxx Output Files"

The EAMxx output file that QuickView has been tested for
was generated in late 2024. As EAMxx further evolves and its output
format changes, QuickView might need to be updated accordingly.
If the user encounters such a case, we recommend reaching out to our lead developer
[(abhi.yenpure@kitware.com)](https://www.kitware.com/abhishek-yenpure/)
or using the [Issue tab on GitHub](https://github.com/ayenpure/QuickView/issues)
to start a discussion.

Overview

The ParaView Reader behind the QuickView GUI detects and categorizes variables based on their dimensions. Variables with dimensions not matching the expected patterns are ignored.


2D variables

QuickView recognizes variables with—and only with—the ^^dimensions time and ncol^^ as 2D variables, i.e., physical quantities varying with latitude and longitude but without a vertical dimension. These are referred to as "surface variables" in the control panel. For these variables, the only required ^^coordinate variable^^ is ^^time^^. Grid information, including latitude and longitude, are obtained from the connecitivity file.

If an area variable with the dimension ncol is also present, this variable is used for calculating the area-weighted horizontal averages displayed in the viewport. If the area variable is not present, then an arithmetic average is calculated and displayed.


3D variables

If a variables has not only time and ncol but also a vertical dimension, QuickView expects the dimension to be named ^^lev^^ for variables defined at layer midpoints and ^^ilev^^ for variables defined at layer interfaces.

For variables defined at layer midpoints, in order for the Slice Selection section of the control panel to work properly, the simulation data file needs to contain a 1D coordinate variable named lev, the values of which are interpreted as pressure in hPa. If lev is not present, QuickView attempts to find two 1D variables, hyam and hybm, of that dimension size, from which QuickView calculates lev using

lev = (hyam * P0) + (hybm * PS0)

where PS0 = 1000 hPa; P0 is read from the data file and set to 1000 hPa if not found.

Similarly, for variables defined at layer interfaces, QuickView looks for either ilev or hyai and hybi for parsing the vertical dimension.


Variable with more dimensions

QuickView currently only visualizes the variables types discussed above. For variables that have extra dimensions in addition to time, ncol, and lev or ilev, for example aggregated tracer arrays or COSP-related variables, support can be provided if there is sufficient interest from the users. Please contact our lead developer (abhi.yenpure@kitware.com) for a discussion.


Missing values

If a variable has an attribute named missing_value or _FillValue, the value is converted to NaN and ignored in the calculation of global averages and for the visualization.