Processamento da linha de drenagem

Question

<\/HEAD>Oi - Estou tendo problemas com a ferramenta de processamento da linha de drenagem. Estou usando a versão 2.0.1.133 no 10. Estou tentando processar um HUC de 8 dígitos usando o NED de 60 m no Alasca. Estou usando um limiar pequeno para iniciar os cursos d'água (~ 1 km2) para que eu possa capturar pequenas nascentes e geralmente corresponder aos cursos d'água confirmados no NHD. Isso resulta em cerca de 5.000 bacias. O GIS trava assim que chego na ferramenta de processamento da linha de drenagem. Se eu clicar em cancelar, a classe de feição drainageline é criada, mas está faltando HydroID, gridID e nextdownID, então não posso continuar com a ferramenta de processamento da bacia adjacente. <\/SPAN> Alguma ideia do que estou fazendo errado?<\/SPAN> Obrigado.<\/SPAN><\/BODY><\/HTML>

MarkBoucher · Answer

I've learned a few things since 2011. You can find tips that will help Arc Hydro work faster and with fewer errors/crashes at the link below my signature.

TobiasOker1 · Answer

I opened the DEM in a new file and everything works just fine. Trick seems to be to get rid of the original layers from which the projected DEM was clipped.

MarkBoucher · Answer

... This results in ~5,000 catchments.  GIS seizes up once I get to the drainage line processing tool.  ...

My Guess is that you are asking for too many catchments, though I'm not aware of any upper limit Arc Hydro has.

Make sure the temp files are cleared out before you start (ApUtilities>Additional Utilities>Clean User's Temp File), and reduce the number of catchments by increasing the initiate streams threshold. (I'm using v10. Not sure if this is in earlier versions. See http://forums.arcgis.com/threads/29202-Solving-Arc-Hydro-Problems-by-deleting-TMP-Files for another way to clear the temp files.).

Also, I've found Arc Hydro will look like it is seized up when it is still processing. Maybe you need to let it go longer before you cancel it. Can you let it run over night?

If at your wits end, start with a fresh mxd and simplify things: small rawdem (takes less time), big initiate streams value (to see if size matters).