Dataset and file structure#
CPCIR and GridSat#
CCIC is derived from two satellite-observations datasets: The GridSat-B1 dataset [KNOAACProgram14] and the NCEP/CPC Merged IR dataset [JJX17]. In the following, these two datasets will be referred to as GridSat and CPCIR, respectively.
CCIC provides estimates for both the GridSat and the CPCIR datasets. The estimates are provided on the same grid as the observations and thus inherit their temporospatial resolution and coverage. With temporospatial resolution of 3h @ 0.07 degree, the GridSat-based data offers lower resolution than the CPCIR data, which has 30 min @ 0.036 degree resolution. However, GridSat is available from 1980, whereas CPCIR only from 2000 (with some retrievals in 1998). The temporospatial coverage and resolution is summarized in table table 1.
Dataset |
Coverage |
Spatial resolution |
Temporal resolution |
---|---|---|---|
GridSat |
1980 - present |
0.07 degree |
3 h |
CPCIR |
2000 - present |
0.04 degree |
30 min |
Data organization#
The CCIC record is organized into results derived from GridSat input data results derived from CPCIR input data. Below that files are organized into folder by year.
record
├── cpcir
│ ├── 2000
│ ├── 2001
┆ ┆
│ ├── 2022
│ └── 2023
└── gridsat
├── 1980
├── 1981
┆
├── 2022
└── 2023
The files follow the naming pattern
ccic_{product}_{YYYYmmddHH}00.zarr
where {product}
is either cpcir
or gridsat
, and YYYYmmddHH
is the timestamp. Each GridSat file contains the retrieval for one timestamp, and each CPCIR file contains the retrieval for two timestamps, the full hour and 30 minutes after the full hour. For each day, GridSat are available at hours HH
= 00, 03, 06, 09, 12, 15, 18, and 21, while CPCIR files are available at hours HH
= 00, 01, …, 22, 23.
To list the files currently available in the CCIC S3 bucket, e.g., for 2020 and CPCIR, you can use:
Python:
import s3fs all_files = s3.ls("chalmerscloudiceclimatology/record/cpcir/2020") first_day = s3.glob("chalmerscloudiceclimatology/record/cpcir/2020/ccic_cpcir_20200101*zarr")
AWS Command Line Interface (terminal):
$ aws --no-sign-request s3 ls s3://chalmerscloudiceclimatology/record/cpcir/2020/
Variables#
The CCIC climate data record provides estimates of the total ice water path (TIWP) and a 2D cloud probability. The data files follow CF conventions. The variables and their meaning are listed in table 2.
Variable name |
Units |
Range |
Description |
---|---|---|---|
|
kg m-2 |
≥ 0 |
Vertically-integrated concentration of frozen hydrometeors |
|
kg m-2 |
≥ 0 |
90% confidence interval for the retrieved TIWP |
|
[0, 1] |
Probability that |
|
|
[0, 1] |
Probability of presence of a cloud anywhere in the atmosphere |
|
|
{0, 1} |
Input pixel was NaN; the retrieval can be a numeric value (inpainted) |
Note: cloud_prob_2d
will be published on a second upload phase.
These variables are gridded on the coordinates from the table below, where the spatial grid and the time resolution are constant for each input product.
Coordinate name |
Units |
Range |
Description |
---|---|---|---|
|
Degrees North |
CPCIR: (-60, +60) |
Latitude |
|
Degrees East |
[-180, +180) |
Longitude |
|
ns |
numpy’s |
Nominal retrieval time |