Laserfiche WebLink
<br />o <br /> <br />60 <br /> <br />100 160 <br />Range. km <br /> <br />200 <br /> <br />rawinsonde data are available only twice per day. <br />Altitudes of the melting level can vary signi- <br />ficantly in 12 hours, especially before and after <br />precipitating weather systems. The numerical <br />model (Eta, Rapid Update Cycle) can indicate the <br />melting level between the rawinsonde observation <br />times and can even identify gradients across the <br />field of view of a radar. However, it is beyond <br />the scope of this effort to modify the SAA code to <br />ingest melting level data. Such is needed to <br />distinguish some bright band effects from other <br />vertical gradient reversals. However, there needs <br />to be caution in the use of melting level data. I <br />have recently observed a case in which there was <br />a layer of air in a precipitating system that was <br />nearly isothermal at a temperature close to 0 oC <br />for a depth of over 1 km. Assuming a bright band <br />of only a few hundreds of meters depth based on <br />derived melting level data would be unreliable in <br />such a situation. Therefore, a bright band <br />warning could not be provided without further <br />study. <br /> <br />KDLH, dens.=O.085 <br />x >=0.1"5;>=1.0.50 <br />c >= 3.0"50 <br /> <br />10 <br /> <br /> x )( <br /> x x x <br />0 x <br />en x <br />~ " x <br />0 <br />"'" <br />12. .. <br /> " <br /> <br />0.1 <br /> <br /> <br />Figure 6e.- The light precipitation data are offset from <br />unity in the opposite direction from the few heavier data <br />for KDLH. <br /> <br />7. ADDITIONAL EFFORTS <br /> <br />7.1 SAA Code Adjustments <br /> <br />In early November 1998, a bug was reported in the NOSPIKES subroutine. Reclamation determined <br />that the bug existed in all previous versions of the SAA. The bug was activated when a pixel had a <br />reflectivity exceeding the upper threshold (40 dBZ) and all 8 adjacent pixels had no detectable echo. <br />Such is a very rare occurrence. The coding was changed to prevent an ALOG 10 of numbers less than or <br />equal to zero. <br /> <br />Coding was further changed to avoid Y2K problems. <br /> <br />Additional coding modifications, shown in the appendices, were created for experimental investigations. <br />Appendix A presents coding modifications to the SAA to make the vertical profile graphs as shown in <br />figures 1 and 7. Appendix B presents a stand-alone program to add many days of SAA output (* .STP <br />files) to show blemishes in the hybrid scan and occultation files. Appendix C presents coding modifi- <br />cations to the SAA to attempt virga removal from the SAA products. The coding of appendices A and C <br />has not been incorporated into the operational versions of the SAA. <br /> <br />7.2 Support Code Development <br /> <br />A separate program (RESIDUAL.FOR, presented in appendix B) was written to ingest many days of <br />daily total (* .STP) files of S accumulations in order to identify residual ground clutter returns or bins <br />overcorrected for clutter. The daily files were partitioned, by the number of radar bins with non-zero <br />accumulations, into three categories of snow fall that might be termed trace-light, moderate, and wide <br /> <br />12 <br />