<style class="ke-style">
[list-style-type] {padding-left:20px;list-style-position:inside}
[list-style-type] li {margin:0}
[list-style-type] li:before, span.ke-list-item-matter {font-family:"sans serif",tahoma,verdana,helvetica}
[list-style-type] li p,[list-style-type] li h1,[list-style-type] li h2,[list-style-type] li h3,[list-style-type] li h4,[list-style-type] li h5,[list-style-type] li div,[list-style-type] li blockquote{display:inline;word-break:break-all}
[list-style-type] li table {display:inline-block;vertical-align:top}
p{margin:0}
td {word-break: break-word}
.default-font-1773847251629{
}
</style><div class="default-font-1773847251629" dir="ltr"><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">Dear MITgcm Community,</span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br></span></p><p style="text-wrap-mode:wrap;"><span><span style="font-size:16px;">In "data" runtime file, I did not set </span></span><span><span style="font-size:16px;">''<span style="font-size:16px;font-family:Arial, Helvetica;">diffKzT</span>" and "<span style="font-size:16px;font-family:Arial, Helvetica;">diffKzS</span>" when run the model. However, I have added values to the "data" in the attachment.<br><br><br>Sincerely,</span></span></p><p style="text-wrap-mode:wrap;"><span><span style="font-size:16px;">Chengming</span></span></p><p style="font-size:14px;"><br></p><blockquote name="replyContent" class="ReferenceQuote" style="font-size:14px;padding-left:5px;margin-left:5px;border-left:2px solid #B6B6B6;margin-right:0px;">-----原始邮件-----<br>
<b>发件人:</b> <span id="rc_from">尚丞明 <22320250156689@stu.xmu.edu.cn></span><br>
<b>发送时间:</b> <span id="rc_senttime">2026-03-18 22:28:15 (星期三)</span><br>
<b>收件人:</b> mitgcm-support@mitgcm.org<br>
<b>主题:</b> LabSea model collapse<br><br><style class="ke-style">
[list-style-type] {padding-left:20px;list-style-position:inside}
[list-style-type] li {margin:0}
[list-style-type] li:before, span.ke-list-item-matter {font-family:"sans serif",tahoma,verdana,helvetica}
[list-style-type] li p,[list-style-type] li h1,[list-style-type] li h2,[list-style-type] li h3,[list-style-type] li h4,[list-style-type] li h5,[list-style-type] li div,[list-style-type] li blockquote{display:inline;word-break:break-all}
[list-style-type] li table {display:inline-block;vertical-align:top}
p{margin:0}
td {word-break: break-word}
.default-font-1773844033603{
font-size:14px;
}
</style><div class="default-font-1773844033603" dir="ltr"><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">Dear MITgcm Community,</span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br style="font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">(Dimitris Menemenlis suggested I move this discussion from GitHub Issue to the mailing list.)</span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br style="font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">I am currently running a high-resolution (3.5 km, <span style="font-family:"Times New Roman", Times;font-size:16px;text-wrap-mode:wrap;">340*468*43</span>) regional model for the Labrador Sea. I am using the EXF and KPP modules. My sea surface forcing field is provided by directly specifying net heat flux (NHF), shortwave radiation, wind stress, and significant fres</span></span><span style="font-size:14px;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">hwater runoff along the coast.</span></span><span style="font-family:"Times New Roman", Times;font-size:12pt;">Specifically,
I placed the runoff separately into the SFLUX channel.</span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><strong style="font-size:12pt;">Problem Description</strong>:<br style="font-size:12pt;">My model runs fine in winter and spring, but crashes in early autumn (September) with a SURF_ADJUSTMENT error (specifically, cg2d solution divergence).</span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br style="font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><strong style="font-size:12pt;">Diagnostic Analysis</strong>:<br style="font-size:12pt;">By examining the output before the crash, I discovered a "heat trapping" issue: </span><span style="font-family:"Times New Roman", Times;font-size:12pt;">In summer, the sea surface temperature (SST) reaches unrealistically high values (approximately 30°C). </span></span><span style="font-family:"Times New Roman", Times;font-size:12pt;">The attached figure displays the 5-day mean temperature distribution: the first row shows
the meridional inflow profile at the eastern boundary, while rows 2 through 4
present the temperature </span><span style="font-family:"Times New Roman", Times;font-size:12pt;">at </span><span style="font-family:"Times New Roman", Times;font-size:12pt;">the
surface, 15 meters, and 60 meters, respectively. Please pay close attention to
the </span><span style="font-family:"Times New Roman", Times;font-size:12pt;">colorbar</span><span style="font-family:"Times New Roman", Times;font-size:12pt;">.</span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">I have attached the crucial configuration and runtime files. In the "output.txt" file, I retained only the last 100 "cg2d" outputs, as there were simply too many of them.</span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:14px;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">MITgcm
version is checkpoint69e.</span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">I eagerly await your suggestions and guidance!</span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br style="font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"><br style="font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;"></span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">Sincerely,</span></span></p><p style="font-size:14px;text-wrap-mode:wrap;"><span style="font-size:12pt;"><span style="font-family:"Times New Roman", Times;font-size:12pt;">Chengming</span></span></p></div></blockquote></div>