Master of Science Thesis –Confidential– Final Report Arctic Offshore Workability “Development and innovative imaging of a probabilistic climatology regarding Arctic sea ice and air temperatures” L. Burg August 22, 2007 Faculty of Civil Engineering · Delft University of Technology Faculty of Civil Engineering Arctic Offshore Workability “Development and innovative imaging of a probabilistic climatology regarding Arctic sea ice and air temperatures” Master of Science Thesis For obtaining the degree of Master of Science in Civil Engineering at Delft University of Technology L. Burg August 22, 2007 Faculty of Civil Engineering · Delft University of Technology Copyright L. Burg &(cid:0)Allseas Engineering, 2007 All rights reserved. Cover image: Sunset at the Chukchi Sea (http://www.esr.org/) Document typesetting: LATEX Delft University Of Technology Department Of Hydraulic Engineering The undersigned hereby certify that they have read and recommend to the Faculty of Civil Engineering for acceptance a thesis entitled “Arctic Offshore Workability” by L. Burg in partial fulfillment of the requirements for the degree of Master of Science. Dated: August 22, 2007 Supervisor: prof. drs. ir. J.K. Vrijling Readers: dr. ir. P.H.A.J.M. van Gelder prof. dr. ir. G.S. Stelling ir. R. Haddorp Abstract Studies underscore the significant reserve potential of the Arctic region: an estimated 25 per- cent of potential oil and gas reserves, onshore and offshore, is believed to be located in the Northern Hemispheres Arctic region. Although these areas already have been under consid- eration for over 30 years, technical difficulties have put a threshold on these developments. Natural conditions like low visibility, extreme colds and ice make it nearly impossible to navigate and therefore restrict workability immensely. This thesis will focus on the occurrences of certain concentrations of sea ice, air tempera- tures and their combined events. Based on these findings a probabilistic approach allows the definition of risk features over defined spatial and temporal domains. Such results will allow a-priori project management for safe Arctic navigation and pipelay operations regarding the conditions under consideration. Different sources of sea ice data are available, but only certain satellites cover the entire Arctic area each day. Unfortunately, resolution is low and their subsequent lack of detecting several features in the ice pack, under which the existence of melt ponds on ice during the summer season, results in an underestimate of sea ice concentrations. Ice charts from the National Ice Center (NIC) overcome this, by using detailed imagery and aerial observations for corrections near the Marginal Ice Zone (MIZ). The NIC charts are therefore especially useful for the specification of sea ice regarding navigational and operational activities and are publicly available through National Snow and Ice Data Center (NSIDC) An extensive analysis has been performed on this dataset with the use of Matlab. The dataset contains ice concentrations and types for a period between 1972-2004 and all areas above latitude 48.52 N. Ice types had to be disregarded due to significant errors in represen- tation, thus reliable (cid:2)indications of ice thickness could not be derived as a result. Ice thickness relations have been investigated, but did not match with uplooking sonar ice draft data. Only light ice conditions are presumed to be allowed and according to rules of thumb, up to 30% ice concentration is considered to be navigable for vessels. During pipelay operations, encountering any ice at all will result in downtime. Efforts have been made to structure and validate the ice concentrations into a reliable, weekly standardized database. Linear regression analyses on ice extent have been performed for different regions and during M.Sc. Thesis v L.Burg vi Abstract different periodes of the year. Most areas show a negative trend over the years 1979 to 2004. Therefore it can carefully be concluded that the analysis on the probability of occurrences for those areas is expected to be on the safe side. It also underlines the variability in seasonal ice extent within the 33-years of observations. Surface air temperature data for Polar Regions are scarce, especially visual representations of the spatial and temporal distributions on a wide scale and offshore. The IABP/POLES dataset is considered the most accurate for surface air temperatures in the Arctic Ocean, based on several drifting buoys and coastal stations. Weekly averaged and minimum air tem- peratures from 1979 to 2004 have been derived on a bilinear interpolated grid, georeferenced with geographic coordinates and matched to the sea ice grid. The probabilities of occurrence for sea ice concentrations above 0% and 30% have been deter- mined by calculating the relative frequency over 33 years. Cross-correlations between surface air temperatures show a lagged correlation of maximally -0.93 after 5 weeks. Seasonal effects, however, seem to dominate with strong autocorrelation up to high lags. Cross-correlation between their anomalies reduce to -0.45 at a lag of 1 week. It is however hard to make a fundamental statement, due to factors such as the influence of ocean-atmospheric coupling. Still, both methods show negative correlations between both parameters with sea ice lagging surface air temperatures, which is as expected. Also open water periods can be evaluated for a specific location. To analyze the effect of air temperatures and ice existence on navi- gability and operability, occurrences of both surface air temperatures and sea ice have been determined. The combined events include: Thresholds for surface air temperatures. (cid:3) – -15 , current structural limit; – -40(cid:4) , designated structural limit; – -5 (cid:4), where icing may start to form. Thresh(cid:4)olds for ice concentrations. (cid:3) – 0% ice concentration, current structural limit; – 30% ice concentration, assumed allowed conditions with ice classification. Results have been visualized on a virtual globe with Google Earth. Layers, which can be controlled over time, display occurrences of all analyzed conditions as well as surface air temperatures. Navigation routes and pipeline tracks can be plotted on top, to dynamically estimate the likelihood of the (combined) events. Not only will these layers be useful for engineering purposes, but any Arctic research related to air temperatures and occurrences of ice. Therefore NSIDC has offered to implement this material online, for both public and scientific use. Within Google Earth and Matlab a combined analysis has been performed on sea ice and air temperature conditions and in the Bering, Chukchi and Beaufort Seas. Based on those figures, decisions can be made on allowed risk and the resulting likelihoods for navigability and operability. Digital elevation models showing bathymetry have been used to indicate the relevance of research on ice scouring for dredging optimization. In short this research has led to the following useful results: Twostructureddatabasescontainingsurfaceairtemperaturesandseaiceconcentrations (cid:3) over the entire Arctic, which has many possibilities in use; L.Burg M.Sc. Thesis Abstract vii Matlab routines that will calculate the probability of single or combined events of (cid:3) assigned limitations in surface air temperatures and sea ice concentrations and possible change in open water length; Dynamic layers within Google Earth for dynamic assessment of Arctic offshore worka- (cid:3) bility regarding sea ice and air temperatures. Implementation hereof by NSIDC reveals the relevance for Arctic research and other purposes; An example of employing the results in a potential future Arctic pipeline project. (cid:3) M.Sc. Thesis L.Burg viii Abstract L.Burg M.Sc. Thesis