A B737-300 ENCOUNTERED A PLASTIC BAG OR SHEET ON TKOF ROLL WHICH COVERED THE CAPT'S PITOT PROBE CAUSING A 40 KT AIRSPD ERROR. ACFT RETURNED TO THE FIELD.

Narrative

ON TKOF OUT OF ZZZ ACFT ENCOUNTERED AN AIRSPD SPLIT BTWN THE CAPT'S AND FO'S AIRSPD INDICATORS. THE SPLIT WAS APPROX 40 KTS. SHORTLY AFTER THE DISCREPANCY WAS ENCOUNTERED IT CORRECTED ITSELF. THE FLC ELECTED TO RETURN TO ZZZ AS A PRECAUTION; EVEN THOUGH AIRSPD INDICATOR FUNCTIONED PROPERLY FOR REMAINDER OF FLT. UPON ARR IN ZZZ; MAINT INSPECTED THE PITOT STATIC PORTS AND FOUND WHAT LOOKED LIKE A PIECE OF PLASTIC MELTED ON THE CAPT'S PITOT PROBE. THEN MAINT REMOVED THE PIECE OF PLASTIC FROM THE SIDE OF THE PITOT PROBE AND DID A BITE TEST OF THE CADC SYS. THE SYS BITE CHK WAS GOOD. A RESEARCH OF HISTORY IN ATA 3400 SHOWED NO PRIOR DISCREPANCIES. IT WAS SURMISED THAT A PIECE OF PLASTIC WAS CAUGHT ON THE PITOT TUBE WHILE THE ACFT WAS EITHER AT THE GATE OR ON TAXI OUT FOR TKOF. IT MELTED ONTO THE PROBE WHEN PITOT HEAT WAS TURNED ON. AFTER ACFT ROTATED AND BEGAN ITS CLB THE FORCE OF THE AIRFLOW PAST THE PROBE MUST HAVE DISLODGED THE PLASTIC; THUS RETURNING THE AIRSPD INDICATION TO NORMAL. LOOSE DEBRIS AT ARPT IS THE SUSPECTED CAUSE.