fxllocet plete fies scuphete dependemse open the statis ground sere wad cyetgh, Buf te fast vbhli remains thet ety system of fall-out fosverting has eo maxizet: upper limit in efficleney defined by thts presision of the wind field forecasting, ‘The discussion herein ip not intended te ienly that CASTLE weather foresasting was por, To the ecntrafy, this service during C/STLE was mech supericr to previous wrk st the FPO. be Since any iaprovesent in the metecrclogisalfield will ultimately reflect itself in improved ecnditious in fall-out forecasting, it seeus imperative thet weather and radsafe pecple ecutinee to sepport and assist each other and to meintain the close mutacl interest in future operations as existed on CASTLE. Rewognining the limiting difficultice verieh constantly hound the westRher forecaster in theyPacific, 4 scams Gr that the logical futures stepe for botk weather and redeefe forecastere we ~ clear, nenaly, (1) Ccertimmed researeh and study of tropical avtocrology, (2) Continued devolopment of new techriques aid refinever®d of the ecrremt fall-cut forecasting systems, particularly those developed oa CLSTLBS a ~ "¢) “Continued evetios ef partiale-sise parameters anc the aftout of thetropopmuse in the fali-cut sechanisn. (h) ta recogiticn of the possitility for error in either the weather cr fall-out forceast, the development of mutual weather/redsafe techniques to "follow through" H-heur with complete post-shet forceasts and observations tn order to obtain the earliest possible warning of the development of post-ehot adveres fakl~cut conditions. a ny 30