HakaiInstitute / hakai-bottle-tool

Python tool use to combine Hakai sample data and CTD data together and output in a NetCDF format to handle by ERDDAP

Hakai Bottle Tools

The hakai bottle tool join together sample and ctd profile data collected by the Hakai Institute and available from the following endpoints within the Hakai API:

{
    "eims/views/output/nutrients",
    "eims/views/output/microbial",
    "eims/views/output/hplc",
    "eims/views/output/poms",
    "eims/views/output/ysi",
    "eims/views/output/chlorophyll",
    "eims/views/output/phytoplankton",
}

The ctd data is retrieved from the API endpoint:

"ctd/views/file/cast/data"

Instalation

You can install the package locally by running for the following command:

pip install git+https://github.com/HakaiInstitute/hakai-bottle-tool.git

You however don't need to install necessarily the package and just use the following jupyter notebook on google colab here.

How To

The hakai-bottle-tool can either be run on the google colab jupyter notebook or, if installed locally, by running the following command:

> python hakai_bottle_tool -station QU39 -time_min 2020-01-01 -time_max 2021-01-01

Method

Each sample type is first groupby site_id, event_pk, line_out_depth and collected time (± 5 minutes) and aggregated by mean (numerical), comma seperated joined strings for strings, count(see _nReplicates), and difference between min and max for numerical values.

All sample type then is joined together by matching site_id, event_pk, line_out_depth and collected time (± 5 minutes).

Once all the sample data available. The corresponding CTD profile data collected over the corresponding time period and station is downloaded and merged to the bottle data by using the following sequence:

1. Bottle data is matched to the nearest CTD profile within 3 hours of the collected time and matched to an exact binned depth if available. If no exact binned depth is available, this bottle will be ignored from this step.
2. Unmatched bottles are then matched to the nearest profile and depth within the depth tolerance.
3. Unmatched bottles are then matched to any CTD collected at taht station within the last day and at the nearest depth within the tolerance
4. Unmatched bottle data left remained unmatched to any CTD data.

A sample is considered within the depth tolerance if the following condition is respected:

where D_ctd and D_bottle corresponds respectively to the CTD and bottle measurements associated depths.