Stephen R. Hudson

Lana Cohen

Markus Kayser

Marion Maturilli

Joo-Hong Kim

Sang-Jong Park

Woosok Moon

Mats A. Granskog

The data has been postprocessed using the stored edt-files.

The data given in [rsData_gridded.nc]( https://api.npolar.no/dataset/216df9b3-e2bd-5111-9c02-fea848d76670/_file/rsData_gridded.nc) is put on a like vertical grid, in order to simplify data handling. The gridding is done with MATLAB ([source code]( https://api.npolar.no/dataset/216df9b3-e2bd-5111-9c02-fea848d76670/_file/n-ice2015_radiosondes_proc.mat)). Missing/errors are marked as "NaN".

Dates/times for each launch is found in the file: [rsData_gridded_dates.txt]( https://api.npolar.no/dataset/216df9b3-e2bd-5111-9c02-fea848d76670/_file/rsData_gridded_dates.txt) where the 6 columns are: year, month, day, hour, minute, second.

The `rsData_gridded.nc` is in NetCDF classic format

```

Dimensions:

"height" = 1201 (0,25,50,...,30000)

"launches" = 258 (1,2,3,...,258)

"dim_launch_date_vector" = 6 (1,2,3,...,6)

"serial_number" = 9 (1,2,3,...,9

Variables:

"height" = height(height)

long name = 'geopotential height'

units = 'm'

Datatype: double

"time" = time(height,launches)

long name = 'time after launch'

units = 's'

Datatype: double

"press" = press(height,launches)

long name = 'pressure'

units = 'hPa'

Datatype: double

"temp" = temp(height,launches)

long name = 'air temperature'

units = 'degree Celsius'

Datatype: double

"rh" = rh(height,launches)

long name = 'relative humidity'

units = '% RH'

Datatype: double

"wd" = wd(height,launches)

long name = 'wind direction'

units = 'degree'

Datatype: double

"ws" = ws(height,launches)

long name = 'wind speed'

units = 'ms^-1'

Datatype: double

"SN" = SN(launches,serial_number)

long name = 'serial number sonde'

Datatype: char

"lt" = lt(launches,dim_launch_date_vector)

long name = 'date of launch'

units = 'yyyy_mm_dd_HH_SS'

Datatype: double

"lat" = lat(launches)

long name = 'latitude'

units = 'dec. degree'

Datatype: double

"lon" = lon(launches)

long name = 'longitude'

units = 'dec .degree'

Datatype: double

"gc_p" = gc_p(launches)

long name = 'GC pressure offset'

units = 'hPa'

Datatype: double

"gc_T" = gc_T(launches)

long name = 'GC temperature offset'

units = 'degree Celsius'

Datatype: double

"gc_RH" = gc_RH(launches)

long name = 'GC humidity offset'

units = '% RH'

Datatype: double

```

#### Issues

During the campaign period only minor problems occurred with the radio sounding:

(1)The biggest problem during processing is the missing ground check values during 57 launches. However, the ground check is just an offset correction and some people argue not to use it in analyzing radiosonde data. (Miloshevich et. al. 2004). I have therefore removed the ground check correction from the profiles and icluded it seperately in the file in the variables "gc_p", "gc_T" and "gc_RH".

(2)Whenever strong winds occurred while filling the balloon on deck, the amount of helium inside the balloon could vary, causing the ascend speed during launch beeing somewhat smaller or faster than 5 ms^-1 in the lowest 1.5-2 km of the troposhere.

(3)A few radiosonde launches where ended very early due to the balloon bursting. The balloons must have gotten damaged during the launch procedure.

CEOS IDN DIF

9.8.4