Introduction

EC-PeT is a new software for processing turbulence measurements from various different sonic anemometer and gas analyser combinations using the eddy-covariance method. The software is written in Python and offers both:

  • A command-line interface for the experienced user that allows easy automation via scripting or planned tasks

  • A graphical user interface for novice users and easy adjustment of the most-used configuration parameters

Its main engine is mostly a rewrite of the well known EC-PACK software. Hence, results are comparable to this software used in a large number of research programs in the last two decades.

EC-PeT reads Campbell Scientific TOA5 files and stores its results in the same file formats as EC-PACK and EC-frame do. Other file formats will be added in the future.

During processing, data and configuration are stored in a project file that is a simple-structured SQLite3 database. This allows easy export of the results, even if the desired output format is not yet supported by EC-PeT itself (e.g. using sqlitebrowser).

History

Predecessors

The Program EC-PeT is an eddy-correlation software developed from eddy-correlation library EC-PACK and the quality control and quality assessment wrapper EC-frame.

EC-PACK

EC-PACK was written by A. Moene and A. Van Dijk at the Royal Netherlands Meteorological Institute (KNMI), Utrecht University Institute for Marine and Atmospheric Research (IMAU) and Wageningen University group for meteorology and air quality.

EC-PACK calculates fluxes and their errors for every requested interval and the selection of usable values is left to the user upon his needs. It is configured by a number of plain text files as sole user interface and reads data from specially crafted NetCDF files.

EC-frame

EC-frame was developed in the German project Terrestrial Environmental Observatories (TERENO) in an effort to create an automated strategy for quality control and quality assessment (QC/QA) of continuous eddy-covariance measurements [Mauder et al., 2013].

It uses the concept to assign a quality index to the results calculated for each interval, i.e. each result is marked by a flag that can take the following values:

  • 0 = good

  • 1 = suspicious

  • 2 = bad

This scheme was implemented as a pre- and a postprocessor around EC-PACK, called EC-frame, by C. Drüe at University Trier chair for environmental meteorology, together with a slight modification of EC-PACK.

Need for something new

Users, especially novices like students, are getting less and less familiar with working on the command line. In particular when four different executables and a number of required configuration files of different styles have to be dealt with. In consequence, a graphical user interface allowing to customize the most common parts of the configuration was developed.

At the same time, maintaining the code became increasingly tedious as EC-PACK and EC-frame are written in FORTRAN, partly in fixed form and partly in free form and the language standard varies between FORTRAN 77, Fortran 90, Fortran 95, and Fortran 2003.

It turned out to be difficult in particular to parallelize the code to take advantage of modern multi-core processors.

In consequence it was decided to rewrite the entire code in Python 2/3 code using - among others - multiprocessing, numpy and pandas libraries.

Name

The acronym “EC-PeT” was chosen to sound similar to its predecessor EC-PACK and to expand to a full name in Latin, which is:

elaboratio concursuum perturbationum Treverensis

It means “eddy-covariance software from Trier” and is a reference to the Roman roots of Trier, which was founded 17 BC and is the oldest city in Germany.