PyTroll (http://pytroll.org) is a suite of open-source easy-to-use Python packages to facilitate processing and efficient sharing of Earth Observation (EO) satellite data. The PyTroll software is intended for both 24/7 real-time operations as well as research and development. PyTroll grew out of the need to provide a resilient and agile platform that can respond quickly to new user needs and new data sources. PyTroll, being open source, stimulates international collaboration, which is vital with the rapid increase of satellite information availability. The PyTroll software development is strongly user driven and has grown over the past eight years from a collaborative effort between the Danish and Swedish national meteorological services to encompass a worldwide community with active contributors. PyTroll is being used at least operationally in the national meteorological services of Denmark, Norway, Sweden, Finland, Germany, Switzerland, Italy, Estonia, and Latvia. However, given its simplicity, minimal demand on user resources, and community-driven approach, it also encourages and facilitates usage of EO data for individual applications. While PyTroll was originally developed to cater to the needs of the atmospheric remote sensing community, it could be equally useful for land and ocean applications and within hydrology. This article provides an overview of PyTroll, with examples showing the capability of some of the core packages.

Analyzing temporal series of satellite data for regional scale studies demand high accuracy in calibration and precise geo-rectification at higher spatial resolution. The Advanced Very High Resolution Radiometer (AVHRR) sensor aboard the National Oceanic and Atmospheric Administration (NOAA) series of satellites provide daily observations for the last 30 years at a nominal resolution of 1.1 km at nadir. However, complexities due to on-board malfunctions and orbital drifts with the earlier missions hinder the usage of these images at their original resolution. In this study, we developed a new method using multiple open source tools which can read level 1B radiances, apply solar and thermal calibration to the channels, remove bow-tie effects on wider zenith angles, correct for clock drifts on earlier images and perform precise geo-rectification by automated generation and filtering of ground control points using a feature matching technique. The entire workflow is reproducible and extendable to any other geographical location. We developed a time series of brightness temperature maps from AVHRR local area coverage images covering the sub alpine lakes of Northern Italy at 1 km resolution (1986–2014; 28 years). For the validation of derived brightness temperatures, we extracted Lake Surface Water Temperature (LSWT) for Lake Garda in Northern Italy and performed inter-platform (NOAA-x vs. NOAA-y) and cross-platform (NOAA-x vs. MODIS/ATSR/AATSR) comparisons. The MAE calculated over available same day observations between the pairs—NOAA-12/14, NOAA-17/18 and NOAA-18/19 are 1.18 K, 0.67 K, 0.35 K, respectively. Similarly, for cross-platform pairs, the MAE varied between 0.5 to 1.5 K. The validation of LSWT from various NOAA instruments with in-situ data shows high accuracy with mean R² and RMSE of 0.97 and 0.91 K respectively.

Atmospheric interaction distorts the surface signal received by a space-borne instrument. Images derived from visible channels appear often too bright and with reduced contrast. This hampers the use of RGB imagery otherwise useful in ocean color applications and in forecasting or operational disaster monitoring, for example forest fires. In order to correct for the dominant source of atmospheric noise, a simple, fast and flexible algorithm has been developed. The algorithm is implemented in Python and freely available in PySpectral which is part of the PyTroll family of open source packages, allowing easy access to powerful real-time image-processing tools. Pre-calculated look-up tables of top of atmosphere reflectance are derived by off-line calculations with RTM DISORT as part of the LibRadtran package. The approach is independent of platform and sensor bands, and allows it to be applied to any band in the visible spectral range. Due to the use of standard atmospheric profiles and standard aerosol loads, it is possible just to reduce the background disturbance. Thus signals from excess aerosols become more discernible. Examples of uncorrected and corrected satellite images demonstrate that this flexible real-time algorithm is a useful tool for atmospheric correction.