oloapinivad/miles

Mid-Latitude Evaluation System

miles's Issues

Possible bug in daily anomalies from running mean

daily.anom.run.mean5() may have a bug induced by the presence of non-continuous season across the end of the year (i.e. winter). Need to check if the seasonal cycle is built starting from the first of January or if we need to remove the half-season and the 29th of Febrauary.

Custom EOFs does not support cross-dateline regions

This is a known bug.
When setting cross-dateline custom EOF region the code will crash since.

Using ERA5 DJFM -> Error in repcheck[, 3] : subscript out of bounds

I'm trying to compare blocking indices to a none standard climatology (DJFM ERA5 1979-2019) and came across the following issue:

###########################################################
MiLES is running on:
ERA5 1979 2019
Active options are:
block figures doforceanl
###########################################################
Full filename is: /p/largedata/hhb19/jstreffi/runtime/oifsamip/MILES/data/zg50000//ERA5/zg50000_ERA5_fullfile.nc
50000
zg50000 NetCDF data seems there, avoid universal_prepare.sh
real 0m0.002s
user 0m0.002s
sys 0m0.000s
Working on DJFM season
List of 6
$ field : num [1:288, 1:144, 1:4971] 51416 51415 51414 51413 51413 ...
$ lon : num [1:288] -179 -178 -176 -175 -174 ...
$ lat : num [1:144(1d)] -89.4 -88.1 -86.9 -85.6 -84.4 ...
$ plev : num [1(1d)] 50000
$ time : 'PCICt' num [1:4971(1d)] 1979-01-01 09:00:00 1979-01-02 09:00:00 1979-01-03 09:00:00 1979-01-04 09:00:00 ...
..- attr(, "cal")= chr "proleptic_gregorian"
..- attr(, "tzone")= chr "GMT"
..- attr(*, "units")= chr "secs"
$ var_units: chr "m2 s-2"
NULL
[1] "Time Array Built"
[1] "Length: 4971"
[1] "From 1979-01-01 09:00:00 to 2019-12-31 09:00:00"
[1] "xcritical: 2.5"
[1] "xreso: 1.25"
[1] "ycritical: 2.5"
[1] "yreso: 1.25"
[1] "Tibaldi and Molteni (1990) index..."
[1] "D'Andrea et al. (1998) index..."
[1] "--------------------------------------------------"
[1] "Blocking indices..."
[1] "gradients"
[1] "anomalies and threshold..."
[1] "Computing D12 instantaneous index ..."
[1] "Computing ExtraD12 instantaneous index ..."
[1] "Computing S04 instantaneous index ..."
[1] "Diagnostics..."
[1] "Creating abind dataset ..."
[1] "Computing Rossby Wave Breaking..."
[1] "dim(repcheck): 4078745" "dim(repcheck): 2"
[1] "dim(ii): "
[1] "steprwb: 6"
Error in repcheck[, 3] : subscript out of bounds
Calls: miles.block.multiple -> cbind
Execution halted

This seems to stem from

MiLES/script/block_multiple_fast.R

Line 209 in 887c128

rwb_west <- new_field[cbind(ii - steprwb, repcheck[, 2] - steprwb, repcheck[, 3])]

It would appear that repcheck is missing one of it's dimension but it's not clear to me where the root cause of the problem lies. I somewhat suspect that it may be with the input file I provide:

cdo sinfo zg50000_ERA5_fullfile.nc
File format : NetCDF2
-1 : Institut Source T Steptype Levels Num Points Num Dtype : Parameter ID
1 : ECMWF unknown v instant 1 1 41472 1 F32 : 129.128
Grid coordinates :
1 : lonlat : points=41472 (288x144)
lon : 0 to 358.75 by 1.25 degrees_east circular
lat : -89.375 to 89.375 by 1.25 degrees_north
Vertical coordinates :
1 : pressure : levels=1
plev : 50000 Pa
Time coordinate : 14975 steps
RefTime = 1979-01-01 00:00:00 Units = hours Calendar = proleptic_gregorian Bounds = true
YYYY-MM-DD hh:mm:ss YYYY-MM-DD hh:mm:ss YYYY-MM-DD hh:mm:ss YYYY-MM-DD hh:mm:ss
1979-01-01 09:00:00 1979-01-02 09:00:00 1979-01-03 09:00:00 1979-01-04 09:00:00
1979-01-05 09:00:00 1979-01-06 09:00:00 1979-01-07 09:00:00 1979-01-08 09:00:00
1979-01-09 09:00:00 1979-01-10 09:00:00 1979-01-11 09:00:00 1979-01-12 09:00:00
1979-01-13 09:00:00 1979-01-14 09:00:00 1979-01-15 09:00:00 1979-01-16 09:00:00
1979-01-17 09:00:00 1979-01-18 09:00:00 1979-01-19 09:00:00 1979-01-20 09:00:00
1979-01-21 09:00:00 1979-01-22 09:00:00 1979-01-23 09:00:00 1979-01-24 09:00:00
1979-01-25 09:00:00 1979-01-26 09:00:00 1979-01-27 09:00:00 1979-01-28 09:00:00
1979-01-29 09:00:00 1979-01-30 09:00:00 1979-01-31 09:00:00 1979-02-01 09:00:00
1979-02-02 09:00:00 1979-02-03 09:00:00 1979-02-04 09:00:00 1979-02-05 09:00:00
1979-02-06 09:00:00 1979-02-07 09:00:00 1979-02-08 09:00:00 1979-02-09 09:00:00
1979-02-10 09:00:00 1979-02-11 09:00:00 1979-02-12 09:00:00 1979-02-13 09:00:00
1979-02-14 09:00:00 1979-02-15 09:00:00 1979-02-16 09:00:00 1979-02-17 09:00:00
1979-02-18 09:00:00 1979-02-19 09:00:00 1979-02-20 09:00:00 1979-02-21 09:00:00
1979-02-22 09:00:00 1979-02-23 09:00:00 1979-02-24 09:00:00 1979-02-25 09:00:00
1979-02-26 09:00:00 1979-02-27 09:00:00 1979-02-28 09:00:00 1979-03-01 09:00:00
................................................................................
................................................................................
................................................................................
......
2019-11-03 09:00:00 2019-11-04 09:00:00 2019-11-05 09:00:00 2019-11-06 09:00:00
2019-11-07 09:00:00 2019-11-08 09:00:00 2019-11-09 09:00:00 2019-11-10 09:00:00
2019-11-11 09:00:00 2019-11-12 09:00:00 2019-11-13 09:00:00 2019-11-14 09:00:00
2019-11-15 09:00:00 2019-11-16 09:00:00 2019-11-17 09:00:00 2019-11-18 09:00:00
2019-11-19 09:00:00 2019-11-20 09:00:00 2019-11-21 09:00:00 2019-11-22 09:00:00
2019-11-23 09:00:00 2019-11-24 09:00:00 2019-11-25 09:00:00 2019-11-26 09:00:00
2019-11-27 09:00:00 2019-11-28 09:00:00 2019-11-29 09:00:00 2019-11-30 09:00:00
2019-12-01 09:00:00 2019-12-02 09:00:00 2019-12-03 09:00:00 2019-12-04 09:00:00
2019-12-05 09:00:00 2019-12-06 09:00:00 2019-12-07 09:00:00 2019-12-08 09:00:00
2019-12-09 09:00:00 2019-12-10 09:00:00 2019-12-11 09:00:00 2019-12-12 09:00:00
2019-12-13 09:00:00 2019-12-14 09:00:00 2019-12-15 09:00:00 2019-12-16 09:00:00
2019-12-17 09:00:00 2019-12-18 09:00:00 2019-12-19 09:00:00 2019-12-20 09:00:00
2019-12-21 09:00:00 2019-12-22 09:00:00 2019-12-23 09:00:00 2019-12-24 09:00:00
2019-12-25 09:00:00 2019-12-26 09:00:00 2019-12-27 09:00:00 2019-12-28 09:00:00
2019-12-29 09:00:00 2019-12-30 09:00:00 2019-12-31 09:00:00
cdo sinfo: Processed 1 variable over 14975 timesteps [0.59s 40MB]

Do you see anything obviously wrong with what I've done here?

Best regards,
Jan

Suggestion of adding `abind` among required packages

Dear Paolo,

I would like to suggest to include the abind package among the required packages because even if the installpack.R file gives the message "All packages are present..." the execution fails due to the missing package abind. After installing that package, everything worked fine.

I should mention that I installed the required packages manually and perhaps installing them through your installation script would have installed the abind package as well.

Thanks.

Best,
Alessandro

issue with time_bnds in netcdf file

I noticed that an error is produced in the function 'ncdf.opener.time' from basis_functions.R when running on netcdf files that contain variable time_bnds

on line:
for (axis in naxis) {print(axis); assign(axis,ncvar_get(a,axis))}

I fixed the error by using nco to strip the time_bnds from my netcdf file first:
ncks -C -O -x -v time_bnds iflile ofile

However, since time_bnds are common in many files ..... it might be better to modify the naxis loop in R