Pollen lidar observations in Warsaw (2013-2021)

The dataset covers the period of 2013-2021 regarding the identified pollination cases observed with the multiwavelength aerosol polarization Raman lidar (Polly^XT-UW) over the EARLINET-ACTRIS site in Warsaw, Poland (52.21° N, 20.98° E, 112 m a.s.l).

Datasets were processed using resources of the Poland-AOD Data Server (http://www.polandaod.pl/). The cases of pollination events were identified on the basis of searching for a bell-like shape in 2-dim (time [UTC], Altitude [m a.s.l.]) quick-looks of the lidar range corrected elastic signals (Pr²) at 532 nm channels (total and polarized) form the Polly^XT-UW lidar far-range transceiver. The background removal process was applied on each quick-look by selecting a rectangular sample of surrounding pixels (dimension of approximately 1 km over 24 h at high altitudes 7-12 km). For every identified pollination event with removed background, the values of volume depolarization ratio (VDR) lower than the background average were multiplied by 1.25 (arbitrarily selected; sensitivity study showed negligible effect of this assumption up to 20% differences).

Datasets overview:

1. categories.txt - represents dates split into three categories of lidar observations. The aim of categories is to separate particular days of lidar observations into bins with similar conditions in terms of: measurement length, cloudiness, pollination bell shape occurrence and presence of other aerosol events. Distinguished three categories, where category 1 is the least, and 3 the most restrictive (refer to section Datasets Specification below).
2. vdr_raw.zip - volume depolarization ratio (VDR) dataset based on lidar quick-looks from category 3 in categories.txt. Separate file represents matrix of VDRs obtained from framing of a bell-like shape and removing the background from it during each day in category 3. Pixels of the bell, based on their pixel-color levels, were converted into VDR from the scale given for each quick-look.
3. vdr_mean_with_dispersions.txt - represents mean and dispersion measurements based on the values retrieved in vdr_raw.zip, file consists of: mean, standard deviation, 25th percentile, 50th percentile, 75th percentile, 90th percentile.

Datasets specification:

1. categories.txt:

• file structure:

1. header - file metadata (file content information, measurement location, label naming convention, categories description).
2. dataset - lidar's quick-look dates classified into predefined categories and split into columns. If date was not classified into next category is marked as "None".

• column by index representation (index. label name - description):

1. DATE_CATEGORY_1_532 - dates in which following conditions were fulfilled (for lidar quick-look at 532 nm polarized channel): recorded > 20h lidar observations; observed low clouds (< 2km) for less than 1h; observed clouds < 4km for less than 2.5h of time.
2. DATE_CATEGORY_2_532 - dates in which following conditions were fulfilled (for lidar quick-look at 532 nm polarized channel): number of days in category 1for which the bell-like shape in depolarization plots is present.
3. DATE_CATEGORY_3_532 - dates in which following conditions were fulfilled (for lidar quick-look at 532 nm polarized channel): number of days in category 2 for which no aerosol layers discernible above boundary layer.

2. vdr_raw.zip:

• general structure - each file represents one pollination event on which bell-like shape was identified (category 3 from categories.txt).
• files count: 54.
• files type: txt.
• filenames format: "YYYYMMDDLidarPol532nm.txt". Example: "20131024LidarPol532nm.txt".
• dataset type: float, None.
• file structure:

1. header - file metadata (file content information, measurement location, source, value representation, axes information, date, lidar channel, bell coordinates in pixels on lidar quick look).
2. dataset - bell-like shape with background removed as a matrix of image in terms of VDR.

3. vdr_mean_with_dispersions.txt:

• identified cases: 54 (category 3 from categories.txt), based on vdr_raw.zip.
• file structure:

1. header - file metadata (file content information, measurement location, label naming convention).
2. dataset - split into columns pollination event date with measurements of: mean, standard deviation, 25th percentile, 50th percentile, 75th percentile, 90th percentile. If background was removed in particular place VDR in that place is marked as "None".

• column by index representation (index. label name - description):

1. DATE - processed quick-look date in format: "YYYY-MM-DD".
2. MEAN_532 - raw VDR average at polarized 532 nm channel.
3. ST_DEV_532 - raw VDR standard deviation at polarized 532 nm channel.
4. Q25_532 - raw VDR 25th percentile at polarized 532 nm channel.
5. Q50_532 - raw VDR 50th percentile at polarized 532 nm channel.
6. Q75_532 - raw VDR 75th percentile at polarized 532 nm channel.
7. Q90_532 - raw VDR 90th percentile at polarized 532 nm channel.

General information:

• Files encoding: Windows (lines: \r\n), UTF-8.
• Decimal separator: dot (.).
• Column separator: tab (\t).
• Header denoted as: hashtag (#).
• No data denoted as: (None).
• Date format represented by ISO 8601.

ATTENTION:

We offer a free access to this dataset. The user is however encouraged to share the information on the data use with the Remote Sensing Laboratory by sending an e-mail to rslab@fuw.edu.pl.

In the case this dataset is used for a scientific communication (publication, conference contribution, thesis) we would kindly ask to acknowledge this data provision by adding the following statement in Acknowledgments: "We acknowledge the data originators A.Tomczak and I.S.Stachlewska for the quality-assurance, evaluation, and provision of data sets of the Remote Sensing Laboratory at the Faculty of Physics of the University of Warsaw, Poland."