Svalbard

Your location: OSI SAF HL / Validation / Svalbard

Sea Ice Validation:
• Greenland
• Svalbard
• Southern Hemisphere
• LowRes Ice Drift

The Sea Ice service at met.no produce daily ice charts covering the Fram Strait to the Barents Sea with main emphasize on the areas around Svalbard. Areas where independent information (manual inspection of SAR, MODIS and AVHRR) are utilized are marked by the ice service. These are then collocated with the OSI SAF ice product. The focus is on areas close to the ice edge. Example of a used area is given in the figure below. To the left (1) a daily Sea Ice service product is show, to the right (2) a polygon with validation data from the same product is shown. Click at the image to enlarge.

Below tables with results from comparing the OSI SAF ice concentration and ice edge products are given. These validation resuls are representing areas close to the ice edge. Areas further into the Artic ice have not been included. Including these areas would have increased the validation "score". The validation results show the relative performance in the areas close to the ice edge and for example illustrates the problems finding the ice edge during late summer.

Validation results for Sea Ice Concentration product

OSI SAF Sea Ice Concentration product compared with met.no Ice Service (OSISAF - Ice service)
Source: oper , Product area: nh

Year Month Match +/- 10% (%) Match +/- 20% (%) Bias Std dev Num compared

2016 11 91.07 95.60 -2.27 8.04 54264

2016 12 84.82 91.60 -3.84 10.54 290128

2017 01 82.52 91.07 -4.15 10.44 214959

2017 02 81.43 91.22 -3.79 10.23 189826

2017 03 85.26 93.64 -2.93 9.09 333127

2017 04 87.33 94.13 -2.28 9.19 334291

2017 05 82.23 90.69 -3.56 10.79 235748

2017 06 78.16 85.94 -5.66 13.19 318795

Validation table updated at: 2017.06.26 16:12 UTC

Validation results for Sea Ice Edge product

OSI SAF Sea Ice Edge product compared with met.no Ice Service (OSISAF - Ice service)
Source: oper , Product area: nh

Year Month Match (%) OSI SAF underestimates OSI SAF overestimates Mean edge difference (km) Num compared

2016 11 98.80 0.70 0.50 10.71 57375

2016 12 97.48 1.79 0.73 13.95 305436

2017 01 97.22 1.98 0.79 14.39 226386

2017 02 97.46 1.44 1.10 10.76 201022

2017 03 98.01 1.13 0.87 12.95 354146

2017 04 98.16 1.14 0.70 13.62 358311

2017 05 97.73 1.54 0.73 16.40 252766

2017 06 96.63 2.63 0.74 28.18 349523

Validation table updated at: 2017.06.26 16:12 UTC

Updated June 14. 2017 - Webmaster Steinar Eastwood