Set up the data

Now that your computer has the required programms for running pamflow you can focus on The Guaviare Project. In this section you will download and get familiar with the data collected on the field. These input files are not only part of The Guaviare Project but are also the data that pamflow requires in any other passive acoustic monitoring project.

Summary:

1. Download audio recordings

The audio recordings you’ll need for this tutorial can be found on Zenodo. This sample data is provided to show you how to use pamflow. If you plan to use these recordings for other purposes, please get in touch and make sure to give proper attribution.

2. Audio Root Directory

During The Guaviare Project 4 passive acoustic sensors where installed for 5 days. The sensors where programmed for recording one minute every 30 minutes so, if everything went well, for each day and each installed sensor 48 files were collected for a total number of 960 one-minute recordings.

The resulting audio files are stored in an external disk that’s been delivered to you by field researchers. The audios are organized as shown bellow

/guaviare_project_external_disk/pam_data_guaviare/
├── MC-002/
│   ├── MC-002_20240229_000000.WAV
│   ├── MC-002_20240229_003000.WAV
│   ├── MC-002_20240229_010000.WAV
│   ├── MC-002_20240229_013000.WAV
│   └── MC-002_20240229_020000.WAV
├── MC-007/ 
├── MC-009/  
└── MC-013/  

Each of the 4 subfolders corresponds to one of the installed sensors and stores the 240 one-minute audio files collected by the sensor. These are the audio files from which pamflow will help you extract metadata, perform quality checks, detect species and extract audio segments. This folder containing all the audio files per sensor is called audio_root_directory.

⚠️ Warning: Ensure the file names of the audio files meets the format above. When working with your own audio data, the audio files need to be named following the nomenclature: {Sensor name}{date}{time}.WAV pamflow will ignore files named after a different structure.

3. Field deployment

Field researchers installed the acoustic sensors and took notes on everything important regarding the installation: coordinates of the site, date and time of installation, sensor characteristics and ecological traits of the deployment site.

These notes were handed out to you along with the recordings in a format called field_deployments_sheet. This is a .xlsx file with one row per installed sensor having all the previously mentioned data regarding the installation of the sensor.

4. Target species

Even though there are many bird species at the monitoring site, the community is only interested in a few of them considered relevant for conservation. Along with the devices_root_directory and the field_deployments_sheet, you were given the list of relevant species for the project, namely, the target_species, which is a .csv file with only one column (scientificName) and one row per each one of the species considered important for this particular project.

Scientific Name
Amazona farinosa
Cyanocorax violaceus
Pitangus sulphuratus
Ramphastos tucanus

These are the three only inputs required to run pamflow. Now that you understand what they are and their structure let’s move on with next section to learn how to get pamflow to read them.