eBird2ABAP 0.2.1

Create ABAP card from eBird data

eBird to ABAP

View on GitHub

The aims of this python package is to produce a dataset of ABAP full protocol card equivalent from the eBird EBD dataset.

The overview process is to find and combine eBird checklists that satisfy the full protocol requirements: same observers, within a pentad, at least 2hr of reporting all species spread over 5 days.

This process does not address "ad hoc" data (e.g., incomplete data in eBird).

Getting started

Install the package from Github

pip install eBird2ABAP

Download EBD:

wget https://ebird.org/data/download?p=prepackaged/ebd_AFR_relJul-2024.tar

Run the function

from eBird2ABAP import *
ebird2abap("data/eBird/ebd_AFR_relJul-2024/ebd_AFR_relJul-2024.txt.gz")

Process

1. Construct the list of valid cards
   1. Group raw EBD data to checklist level information (merge shared checklist)
   2. Filter checklists which could make a valid card
      1. Keep only complete checklists
      2. Keep only checklists with Historical, Stationary, Traveling, Incidental protocol
      3. Keep only checklists within the pentad, that is,
         1. Exclude checklists with historical protocol that don't have a distance.
         2. Exclude checklists with distance greater than the distance from center of checklist to closest pentad limit (accept some overlap with a correction factor).
      4. Keep only checklists with duration greated than 0.
   3. Group checklists by date (named checkday later on)
   4. Group checklists into pentad_observer group so that we only have to loop through the date to find valid card
   5. Preliminary filter to eliminate all pentad_observer for which the sum over the entire period does not lead to 2h
   6. For each remaining pentad_observer, apply the function checkday_pentad_observer(), which,
      1. Compute the temporal distance between all checkday and check if they are within 5 days.
      2. Loop through all checkday,
         1. Compute the total duration of all checkdays within temporal distance
            1. If valid, create the card_id and apply it to all checkday. Iterate to the first next checkday that was not within temporal distance
            2. If invalid, iterate to the next checkday
2. Create the card data
   1. For each valid card, aggregate all checklists which are (1) within pentad, (2) same observer and (3) day within the 5 day period. This include more checklists than used to construct the list of valid cards
3. Add species level information to cards
   1. Add sequence information based on first occurance on checklist.
4. Export in JSON

Matching entry

Card level

	Example	eBird EBD	Comments
Protocol	"F"	"F"	Only full card are considered in this conversion process
ObserverEmail	"[e-mail]"	"[e-mail]"	Using KBM email adress for now
CardNo	"0910c0725_050642_20230815"	Pentad_ObserverNo_StartDate	build from Prend, ObserverNo and start date.
StartDate	"2023-08-14"	min(OBSERVATIONDATE)
EndDate	"2023-08-17"	max(OBSERVATIONDATE)
StartTime	"06:43"	min(OBSERVATIONDATETIME)
Pentad	"0910c0725"	card_pentad
ObserverNo	"050642"	"22829"	eBird account number is 22829 on ABAP.
TotalHours	"1:57"	sum(DURATIONMINUTES)	sum of the durations of all checklists
Hour1	14	""	We don't record sequence in eBird
Hour2	23	""
Hour3	0	""
Hour4	0	""
Hour5	0	""
Hour6	0	""
Hour7	0	""
Hour8	0	""
Hour9	0	""
Hour10	0	""
TotalSpp	23	length(sp_list)
InclNight	"0"	"0"	This conversion tool does not adress the issue of nocturnal
AllHabitats	"0"	"0"	This conversion tool does not quantify the use of all habitat.
		Checklists	List of checklists used
		TotalDistance	sum of distances of all checklists
		ObserverNoEbird	observer ID from eBird

Record level

	Example	eBird EBD	Comments
Sequence	1	i	Taxonomic order
Latitude	9.0910404	checklist_latitude	Not recorded
Longitude	7.4309485	checklist_longitude	Not recorded
Altitude	469.8	""	Not recorded
CardNo	"0910c0725_050642_20230815"	Pentad_ObserverNo_StartDate	Same as the card to which the record belong to
Spp	314	ADU	ADU number match based on https://github.com/A-Rocha-Kenya/Birds-of-Kenya
Accuracy	35.340999603271	""	Not recorded
SightingTime	"2023-08-15T05:36:33.834Z"	checklist_start	Use the checklist of the first occurance

Sample ouput

[
   {
      "Protocol":"F",
      "ObserverEmail":"[e-mail]",
      "CardNo":"0500b0220_r1034990_20180213",
      "StartDate":"2018-02-13",
      "EndDate":"2018-02-15",
      "StartTime":"15:08",
      "Pentad":"0500b0220",
      "ObserverNo":"22829",
      "TotalHours":2.0,
      "Hour1":"",
      "Hour2":"",
      "Hour3":"",
      "Hour4":"",
      "Hour5":"",
      "Hour6":"",
      "Hour7":"",
      "Hour8":"",
      "Hour9":"",
      "Hour10":"",
      "TotalSpp":38,
      "InclNight":"0",
      "AllHabitats":"0",
      "Checklists":[
         "S43361123",
         "S43361118"
      ],
      "TotalDistance":0.0,
      "ObserverNoEbird":"obsr1034990",
      "records":[
         {
         "Sequence":1,
         "Latitude":4.9640506,
         "Longitude":-2.40952,
         "Altitude":"",
         "CardNo":"0500b0220_r1034990_20180213",
         "Spp":1338,
         "SourceSpp":,
         "Accuracy":"",
         "SightingTime":"2018-02-13T15:08:00Z"
         },
         ...
      ]
   },
   ...
]

Discussion

• Support Ad-hoc?

What eBird/eBird user can do to improve/maximize the number of card?

• Both protocol need user to do complete list (and not add-hoc or incomplete).
• Try to encourage birder to avoid using Historical protocol
• Use the track to determine pentad overlap
• Let the user confirm that a checklit belong to a single pentad rather than a check based on distance
• User input for AllHabitats.