jhollist / nsink Goto Github PK

Try with EPSG codes: 2234

There is an intermittent bug on generating the static maps. Happened ~3 times on my tests. Throws this error:

Error in CPL_geos_op2(op, st_geometry(x), st_geometry(y)) : 
  Evaluation error: IllegalArgumentException: point array must contain 0 or >1 elements.

I think it is due to a starting location for a flowpath that results in a LINESTRING with only one point. Not sure if this occurs becuase the flow direction creates a single point flow path or if the point is directly on an existing flowline. Also not sure if this is occurring in the flowpath generation or in the flowpath summary. Been hard to recreate and find true culprit.

Unsatisfying fix for now is to re-run nsink_generate_static_maps as the bug is rare and unlikely to show itself twice...

Part of flowpaths were missing due to the disconnected NHD Plus streams. See example for East River-Frontal Guilford Harbor (huc12 = 011000040203).

• nsink_build
• nsink_calc_removal
• nsink_generate_flowpath
• nsink_generate_static_maps
• nsink_get_data
• nsink_prep_data
• nsink_summarize_flowpath

Package level documentation needs to be filled out. @qkellogg do you think you could take a stab at that. The file that needs work is https://github.com/jhollist/nsink/blob/master/R/nsink.R

Do we have info for off network streams and lakes (e.g. kettle ponds) to calculate removal?

Need to figure out a way to include in flowpath when flowlines are not on network.

Jones River good example of these things happening.

Currently these aren't constrained. Ping Art and Q for thoughts on this.

The check for the HUC assumes a character... If passed as numeric, coerce to character then check

I don't think I need the two methods anymore as I think I have it hard coded for hybrid everywhere. Look into removing this option...

in both nsink_generate_flowpath and nsink_summarize flowpath. Look for other instances and see if I can standardize to just a single method.

Hi:
I get the following error when running the nsink_get_data function:
manhan_download <- nsink_get_data(manhan_huc_id, data_dir = "nsink_manhan_data")
Error in nsink_get_plus_remotepath(rpu[i], "NHDPlusAttributes") :
Neither https://s3.amazonaws.com/edap-nhdplus/NHDPlusV21/Data/NHDPlusNE/NHDPlusV21_NE_01_NHDPlusAttributes_09.7z or https://s3.amazonaws.com/edap-nhdplus/NHDPlusV21/Data/NHDPlusNE/NHDPlus01/NHDPlusV21_NE_01_NHDPlusAttributes_09.7z is a valid URL

I ran the Niantic example in the vignette and received the same error message.

Think the base url in the utils.R code needs to be changed from "https://s3.amazonaws.com/edap-nhdplus/NHDPlusV21/Data/NHDPlus" to "https://edap-ow-data-commons.s3.amazonaws.com/NHDPlusV21/Data/NHDPlus"?

Evidently there are many reaches with 0 residence time in the NHD Plus. Look at these and see if I can calc removal using equations in Kellogg et al.