Operational ice charting in mid-latitudes using Near-Real-Time SAR imagery Sergey Vernyayev Ice Engineer ICEMAN.KZ Carles Debart Project Manager Energy, Environment and Security Yevgeniy Kadranov Ice charting specialist ICEMAN.KZ Anton Sigitov Ice and IT specialist ICEMAN.KZ KSAT / 1 /
KSAT CONTENTS introduction Partner s background Operational ice charting needs in the region Description and peculiarities of region Use of satellite imagery Comparative analysis of imagery (optical and radar) Conclusions / 2 /
KSAT COMPANY introduction BACKGROUND ICEMAN.KZ Provision of ice and metocean support services for operations in ice covered waters Pre-season activities Ice season activities Post-season activities Non-operational activities Planning of operations in ice covered waters Training of staff for several levels of awareness on ice related issues Development, setup and integration of data management systems Set-up and operational monitoring of dataflow Operational ice monitoring, regional/operational ice charts, Quicklooks, heli recces, reporting Operational analysis and forecast of ice events hazardous for operations On-site awareness training of operational personnel Evaluation of ice season and impact on operations Development of recommendations to increase safety and efficiency during ice season. Statistical analysis of ice and metocean data Satellite imagery interpretation In-situ ice measurements Sea ice structure interaction scenarios Concept development and design support of structures in ice / 3 /
KSAT COMPANY introduction BACKGROUND KSAT Kongsberg Satellite Services Green Area - direct downlink to Svalbard 15 minutes OIL SPILL & SEEP DETECTION VESSEL DETECTION Brown Area on board storage and downlink to Svalbard 1 hour MULTIMISSION SAR DATA DELIVERY ICE EDGE MONITORING ICEBERG TRAKING Blue Area extended onboard storage and downlink to Svalbard 1h 30 minutes World leader in providing SAR imagery in NRT and Time Critical services / 4 /
OPERATIONAL ICE CHARTING Area of study Kashagan oil field and surroundings - Northern part of the Caspian Sea North Caspian Sea (Lat: 44-47 N, Long: 48-53 E) Revisit of single satellite up to 3 days Unstable weather Rapid change of ice conditions Cloudy days prevent reliable use of optical imagery / 5 /
OPERATIONAL ICE CHARTING Area of study Kashagan oil field and surroundings - Northern part of the Caspian Sea Peculiarities of ice conditions: Variability from year to year depending on severity in: Ice coverage Ice mobility Ice thickness Large number of deformed ice features (ridges, stamukhi) Navigation Challenges: Shallow water depth (4-8m) Ice ridges, Stamukhi Ice Drift Ice Pressure Water level changes Seal habitat area / 6 /
SATELLITE ACQUISITION PROGRAMS Challenges in mid-latitude areas Daily coverage Not possible to achieve with single satellite sources Not a problem for ice charting in the arctic Becomes a greater problems as we approach the equator KSAT Multi-mission solution Uses all available SAR satellite in NRT Achieves daily coverage even at the equator Redundancy means reliability Supported by free optical sources like Landsat, MODIS and Sentinel-2 / 7 /
SATELLITE ACQUISITION PROGRAMS KSAT in support of operational ice charting Multimission Near Real Time Medium Coarse resolution 24/7 resolution ordering Tailored & flexible products ready <30 minutes images (19-25m) images acquisition (50-100m) to plans use for ice charting teams / 8 /
COMPARATIVE ANALYSIS Case study: extreme ice feature monitoring Ice characterization with SAR Multimission approach Different look and feel based on: sensor type spatial resolution incidence angle Continuity is needed for interpretation State of ice matters (refrozen pancake ice gives high response similar to ridged areas and ridges) /9/
COMPARATIVE ANALYSIS Case study: reference features (Stamuki) Stamukhi Ice debris accumulation, grounded on the seabed and emerging above ice surface Typical ice features with rough surface for the shallow zone of the North- Eastern Caspian sea. Immobile through the season and the most noticeable and easy to find during the visual interpretation of an image Can be traced on various images in the same location over a significant period of the ice season Identification of four significant standing alone targets (hazardous ice features stamukhi*) Ground proving of features was done from helicopter flight photos Tracking algorithm - Comparison of images that: a) don t move while everything else does b) channels through ice cover c) high response points in open water on radar that are known not to be vessels / 10 /
COMPARATIVE ANALYSIS Case study: setup (11th to 26th of February 2016) Date of pictures: 20th Feb 2016 Number of image platforms and modes for comparison: 7 3 Ice Charters giving independent opinion based on experience visually identifying features Similar setup as when in operations / 11 /
COMPARATIVE ANALYSIS Optical sensors Landsat 7 & 8 Easy interpretation of water, ice, ground Separate ice features with sizes of >150m can be traced, but not always reliably Operationally unreliable due to weather and delivery time Good for post-analysis / 12 /
COMPARATIVE ANALYSIS Optical sensors Sentinel-2 Easy interpretation of water, ice, deformed ice, ground etc. Even small features (<50m) can be distinguished easily Operationally unreliable due to weather and delivery time Good for post-analysis / 13 /
COMPARATIVE ANALYSIS Optical sensors MODIS Basic discrimination of water, ice, ground Features of >1-2km can be traced, but not always reliably Operationally unreliable due to weather / 14 /
COMPARATIVE ANALYSIS TABLE Optical sensors Name Type Resolution and Coverage Pros Cons MODIS (terra and aqua) Optical, multispectral Full region of NC Sea 250-500 m, 1-2km Free, Operational - 2 times per day, covers whole region, easy interpretation Weather dependent Not enough resolution for detection of small features Landsat (7,8) Optical, multispectral Full region of NC Sea 30m Free, covers whole region, Good resolution, easy interpretation Weather dependent several per month No NRT delivery available in region Sentinel-2 Optical, multispectral Full region of NC Sea 10,20,60m Free, usually covers part of the region, very good resolution, easy interpretation Weather dependent several per month Big delays in receiving big file size / 15 /
COMPARATIVE ANALYSIS Radar sensors - Radarsat-2 SCNA dual pol (HH+HV) Several modes and polarizations available Requires interpretation Features >100m can be distinguished easily Operationally reliable, No daily coverage RADARSAT-2 ScanSAR Narrow, HH+HV, Data and Products MacDONALD DETTWILER AND ASSOCIATES LTD. (2016) All Rights Reserved. / 16 /
COMPARATIVE ANALYSIS Radar sensors - RISAT-1 CRS circular pol (RH+RV) Circular polarization appears superior in extreme feature detection Requires interpretation Features>100m can be distinguished easily Operationally reliable No daily coverage CRS mode, HH, RISAT-1 (2016) ANTRIX Corporation / 17 /
COMPARATIVE ANALYSIS Radar sensors Sentinel-1 IW mode (VV) Requires interpretation Features>50m can be distinguished easily Operationally reliable, No daily coverage Not possible to task, fixed acquisition plans Image copyright Copernicus data 2016 (ESA) / 18 /
COMPARATIVE ANALYSIS Radar sensors TERRASAR-X ScanSAR Wide (HH) Several modes available Requires interpretation Features >50m can be distinguished easily Operationally reliable, No daily coverage TerraSAR-X 2016 Airbus Defence and Space / Infoterra GmbH. / 19 /
COMPARATIVE ANALYSIS TABLE Radar sensors Name Type Resolution and Coverage Radarsat-2 RISAT-1 Sentinel-1 SAR (HH / HV) SAR (RH / RV) SAR (VV) Full/part of region 25-50,50-100 m Part of region 25-50 m Full/part region 15m Pros Different modes, several polarizations, medium/high resolution Different modes, several polarizations, high resolution Circular polarization Free Great resolution and big coverage Cons 2-3 per week Requires interpretation Heavily utilized prone to conflicts 2-3 per week Requires interpretation 1 per week Not always operational One polarization Requires interpretation TerraSAR-X SAR (HH) Full/part region 19-40 m Different modes, High resolution 2-3 per week Requires interpretation / 20 /
COMPARATIVE ANALYSIS Detection performance over a single feature Optical sensors Radar sensors / 21 /
CONCLUSIONS Analysis outcome 1. Optical imagery Low to medium resolution is cheap/free Easily interpreted Not operationally reliable due to clouds In some cases, big file sizes and slow delivery time 2. SAR imagery Reliable Operational in Near Real Time High resolution (adjustable resolution) Not free, but cost effective Harder to interpret (expertise required) 3. An operational ice charting service needs Daily basis optical and radar, best with Near Real Time delivery Reliability and flexible acquisitions plans Different sensor modes Several polarizations enhance visual interpretation of SAR images / 22 /
Thanks for you attention! Any questions? Carles Debart - carlesd@ksat.no Sergey Vernyayev sergey.vernyayev@iceman.kz