As ice monitoring operations take place in the winters Company has sufficient time to conduct research and development activities. These are mainly targeted to collection of historical databases for various areas in the world where we provide or plan to provide our services, investigation of new data sources on the market of satellite images and development of software to manage data and facilitate ice charting processes during operations.
Remote Sensing – Data Sources
Our company has vast experience in using satellite and weather model data for ice charting purposes. However, this is dynamically developing area of business with new satellite platforms starting up regularly. We are continuously monitoring the market of remote sensing and investigate possibility of operational integration of both SAR and optical images into ice charting operations.
As mentioned above our major focus is on Sentinel data that is freely available and it becomes our specialty. We have gained significant experience in processing the raw files from Copernicus, mastered automation to download images with use of API hub and most importantly we can plan the times and coverage of acquisitions with as much certainty as the uncontrollable system allows. But we are always considering the new options coming out to the market. For example, finnish constellation of SAR satellites ICEYE.
Remote Sensing – Analytics
Through our endevours processing huge anount of remote sensing and metocean data our personnel works to rationalize data floes and develop concepts describing physical processes of interaction between metocean factors and ice behavior. Thus, we create sophisticated data models enabling our forecasting and analytical technics in support of operations. These models have their reflection in data management software functions as well as structurize supporting databases.
To facilitate our forecasting technics, we continuously search for historical data to feed our database. To the date we have compiled probably the fullest collection of freely available satellite images for the Caspian Sea region. Our Archive goes back to 70s and with recent work to digitize aerial surveys of the Soviet times the data goes back to 1927. Categorization and ice thickness distribution based on this data is taking place at the moment to merge into analyzable database.
Data Management Software Development
Our Internal database is built on PregreSQL PostGIS platform with data management tools for geospatial input and analysis implemented in QGIS environment with Python scripts ensuring quality and data continuity.
ICEMAN-DMS QGIS plugin package is developped in house. This Data Management System (DMS) package is the latest generation of ice charting package incorporating multiple functions from input of basic ice parameters based on daily observations and up to fully controlled modelling of ice environment based on weather observations now and historically observed over a region. To the current date ICEMAN-DMS is fully operational for the Caspian Region. It was tested for bugs and ergonomics with multiple rigorous Quality Assuarance limitations to avoid human error on entry of data during multi-year ice data hindcast for the Caspian region. Current Functionale above convininet ergonomic structuring of data input includes:
- Automatic assignment of Stages of Development from WMO Ice Egg based on modelled ice thicknesses and storage of both attributes in enhanced Sigrid-3 file. Freezing Degree Days to Ice Thickness Model for the Caspian region is based on emperic formula taking into account sub-arctic conditions, where effects from sun radiation become significant in the second half of the season. The model was verified to numerous in-situ measurements across the Caspian Region.
- Integrated drift analysis module allows background record by zone of wind-drift statistics for detailed evaluation of ice events over regions of the Sea.
The light version of ICEMAN-DMS is available for free distribution on request for non-commercial use by researchers and students. It has basic capability to provide analysis of ice cover based on remote sensing data. It fits training requirements, but is fully stripped of QA, link to hindcast data and modeling ice behavior that facilitate operational processes during commercial services provision.
ICEMAN-DMS Light is an open code plugin for QGIS 2.18 and needs our assistance at initial installation as well as basic introduction to operational principles. Current development is ongoing for upgrading the plugin to QGIS3 and introducing distribution through QGIS repository. for both versions of the plugin the only Copyright restriction for using the code is giving the credit to ©ICEMAN.KZ, when distributing or publishing data generated with use of the plugin.
Driftica is a separate module developped to analyse ice displacements in a region by tracking floes and automatically vectorizing displacements. The QGIS plugin was first created to analyze ice drift behavior in the Caspian region. The outcome of the research was published on POAC2017 in Busan, Korea. The article can be found in POAC repository to see what sort of results can be achieved with the use of the tool.
It is now fully integrated into ICEMAN-DMS as a part of functions there to facilitate daily ice charting operations and generate daily statistics for analysis of ice drift events over zones of a region or the whole region if needed.
Current version of the plugin is fully develpped for Caspian Sea and Ob Estuary regions taking into account specific conditions such as bathymetry and coastline configuration that affects seasonal behavior of ice cover.
Ice Watch standalone package was developped for remote sites without trained ice observers for manual input of visual ice related observations. It has rigid control on entry of ice parameters ensuring minimal human error at entry of data as well as ergonomic list of instructions on ice observations. As it is applied in the areas with limited internet connection, the package is regularly producing xml exerpt of collected information to send email for further integration into geodatabases to be used in ice charting processes for verification of modelled data and remote sensing observations. It was initially developped for agip kco and utilized later by NCOC in their daily operations by fleet’s Bridge crew and Seal Observers to enable efficient ice data exchange between all vessels en route and Ice Charting Office.
Artificial Intelligence and Machine Learning
Huge amount of remote sensing data that became available during the last years with a lot of data providers and our own efforts to generate historical archives in the Caspian sea have enabled us to experiment with Artificial Intelligence and Machine Learning. The first analytical tools based on this technology is already being incorporated into our Data Management System and await verification and further developement during routine operations.