On December 31, 1996, about 1700 Alaska standard time, a wheel/ski equipped Cessna 180, N7066, crashed during the landing approach to runway 34 at St. Mary's, Alaska. The airplane was being operated as a visual flight rules (VFR) cross-country government flight under Title 14 CFR Part 91 when the accident occurred. The airplane, operated by the Alaska State Troopers, sustained substantial damage. The certificated private pilot, and the sole passenger were not injured. Visual meteorological conditions prevailed. A VFR flight plan was filed. The flight originated at the Bethel airport, Bethel, Alaska, at 1612. The pilot reported that the flight was a return to St. Mary's after transporting a prisoner to Bethel. The wheel skis were retracted. During a straight-in approach to runway 34, the pilot applied carburetor heat about 600 feet above the ground in preparation for landing. The engine began to run rough. The pilot continued with the approach and about 300 feet above the ground, began increasing the throttle. The engine quit and the pilot performed an emergency landing off the left side of the runway. He reported during the descent, the left wing began to stall and he increased the rate of descent and turned left. The airplane touched down on snow covered tundra and the left main landing gear collapsed. The airplane received damage to engine propeller, landing gear, and fuselage. At 1656, an automated weather observation system (AWOS) weather observation at St. Mary's was reporting, in part: Wind, 040 degrees (true) at 5 knots; visibility, 5 stature miles; sky condition, clear; temperature, -2 degrees F; dew point temperature, -8 degrees F; altimeter, 29.65 inHg. The pilot reported observing ice crystals visible in the visual approach slope indicator (VASI) light beam for runway 34. On January 15, 1997, the airplane was examined by the National Transportation Safety Board (NTSB) investigator-in-charge (IIC) at the Alaska State Trooper's maintenance facility, Anchorage, Alaska. The fuel selector was positioned to the "both" position. The throttle, propeller, and mixture controls were all full forward. The carburetor heat control was in the "on" position. Fuel was present in the gascolator and was free of contaminants. The propeller blades remained attached to hub and engine crankshaft. Both propeller blade tips were curled aft and exhibited chordwise scratching and leading edge gouging. The propeller was removed and a short "club" prop was installed. A temporary fuel supply was utilized and the engine was started. The engine ran at idle RPM with the carburetor heat control in the "Hot' and "cold" positions. Full throttle application was not attempted. Federal Aviation Regulations (FAR), aircraft certification standards specified in FAR 23.1093, specify that carburetor heat systems are capable of increasing carburetor air temperature by 90 degrees F. Examination of several carburetor icing probability charts, produced by various studies, revealed no charts of icing probability below zero degrees centigrade. In an NTSB special study, "Carburetor Ice in General Aviation", dated January 19, 1972, the phenomena of carburetor icing was addressed and several issues were discussed. The section, Carburetor Ice Formation and Prevention, states in part:..."Any one or combination of these ice-forming situations may cause loss of power by restriction of induction flow and interference with an appropriate fuel-air ratio. One reason it can be important to use carburetor heat as an anti-icer rather than a deicer lies in the "vicious circle" aspect, especially in fast-forming conditions and when the ice buildup might not be diagnosed at an early stage. An uncorrected carburetor ice condition can mean less power, and thus reduced carburetor heat which may result in the formation of more ice. It is certainly only prudent to guard against a buildup of carburetor ice before deicing capability is lost." In the section of the safety study, "Excessive Use of Carburetor Heat", the study states in part:..."There are exceptions to the rule that carburetor heat application results in lower power. In extremely cold and dry weather, with no icing potential, the use of a little carburetor heat may actually increase power to a small extent because of improved fuel vaporization. This reversal of the usual, however, would not occur in most localities." In the section of the safety study, "Prevention Procedures", the study states in part:...12. Consider that carburetor icing can occur with ambient temperatures as high as 100 degrees F and humidity as low as 50 percent....The possibility of carburetor ice decreases in the range below 32 degrees F. This is because of (a) lessened humidity as the temperature decreases, and (b) at around 15 degrees F, any entrained moisture becomes ice crystals which pass through the induction system harmlessly. It should be remembered that if the intake air does contain these ice crystals, carburetor heat might actually cause carburetor icing by melting the crystals and raising the moisture laden air to the carburetor icing temperature range....13. Prior to closed-throttle operation, such as for a descent, apply full heat and leave on throughout throttled sequence. Periodically, open throttle during extended closed throttle operation so that enough engine heat will be produced to prevent icing. Be prepared to remove carburetor heat if go-around is initiated."