Significance of Surface Temperature in the Thermal Infrared Sensing of Sea and Lake Ice

Abstract

ABSTRACT. Therma l infj·ared sensing can provide much information a bout sea ice, and some of th e physica l conditions associated with sea ice suggest that surface tempe ra ture may be a good indi cator of ice thickness. However, steady-sta te heat-Aow calcul ations sugges t that the variable thickn ess of the snow cover a nd its low, variabl e th e rmal conductivity would preclude th e u se of surface tem p e rature a lone as a suitable indi cato r of ice thi c kn ess. Measurements of surface temperature, snow depth, a nd ice thickness suggest that, in an area of rela ti vely uniform ice thickness, surface temperature might be useful as an indi ca tor of snow dept h if some surface d a ta can be obta in ed. RESUME. Much of the past effort in arctic applications of infrared sensing has been directed toward sea -ice reconnaissance (Ander son , 1962 ; Poulin and others, 1962 ; M cLerran , 1965 ;Poulin, 1965 ; K etchum and Wi ttmann, 1966 ; Poulin and Harwood, 1966;Kuhn and others, 1973). These works have shown that r elative ice thicknesses can usually be identified ; new, young, first-year, and multi-year ice can often be delineated as such ; cracks and pressure ridges can be distinguished , and various other ice charac teristics can be determined. Kuhn and others (1973) indicate that it may b e possible to estimate ice thickness with an accuracy of ±o.3 m.The accuracy to which it is necessary to know the thickness of sea or fr esh ice d epends on the type of ice and the operational circumstances involved . Wher eas the captain of a large icebreaker may consider it sufficient to distinguish between first-year or multi-year ice, that of a smaller ship m ay need to know whether or not the ice is thicker than som e critical limit. Other persons may find a difference of a few centimeters to b e significant, even in relatively thick ice. Therefore, it is difficul t to say how accurate a method of ice thickness d etermination ought to be. Although a n accuracy of 5 to 10 % may be d esired , one of 20 to 30% would certainly be useful. It is doubtful whether even the latter accuracy can be a chieved on the basis of surface temperature m easurements alone.Some aspects of sea ice make it an excellen t subject for infrared sensing. For exampl e, its lower surface terminates at a n early constant temperature regardless of ice thickness or geographical distribution ; disturbances in the ice sheet are subject to sharp th ermal contrasts with res pect to the rest of the ice sheet; compared to other terrain material s, the thermal 277

The Analysis of the Near-surface Energy Transfer Environment from Thermal Infrared Imagery

“…Some effort should be expended on the prosecution of an integrated modeling and field-measurement experiment under conditions in which solar radiation is insignificant and thaw zones are present in the ice cover. Attention should also be paid to the effects of variable depth and structure of the snow cover (Poulin, 1975). The goal of this venture should be the designing and testing of a “conservative ice-thickness model”which would permit underestimation of thickness but exclude the possibility of overestimation.…”