From 8a2831b7f24fdc9c2dd19353c8c485a61aaa4889 Mon Sep 17 00:00:00 2001
From: Matt Fisher <matt.fisher@nsidc.org>
Date: Wed, 9 Aug 2023 18:12:16 -0600
Subject: [PATCH] Update the quality checklist from original spreadsheet

---
 doc/contributor/how-to/contribute-layers.md | 32 +++++++++++++++------
 1 file changed, 24 insertions(+), 8 deletions(-)

diff --git a/doc/contributor/how-to/contribute-layers.md b/doc/contributor/how-to/contribute-layers.md
index 1818675d..3fbd3f49 100644
--- a/doc/contributor/how-to/contribute-layers.md
+++ b/doc/contributor/how-to/contribute-layers.md
@@ -96,12 +96,24 @@ A correct QGreenland data pipeline will output data that:
 - Citation populated
 - Layer descriptions populated
 - Layer description excludes data provider
-- Layers'/groups' title follows date standard
-- Layers'/groups' title follows capitalization standard
-- Layers'/groups' title excludes data provider (see notes for exceptions)
+- Layers'/groups' title follows date standard: If a layer is date-specific, the date
+  should be the last part of the title except the resolution, if present. Not in
+  parentheses or otherwise distinguished from the rest of the title. Format is "Month
+  DD YYYY" (all fields optional), e.g.:
+    - `Common Murre Colonies 2010`
+    - `Sea Ice Age September 17-24 2010`
+- Layers'/groups' title follows capitalization standard: Capitalize first letter of the
+  Title. Other than this, only capitalize proper nouns and acronyms. E.g.:
+    - `Thickbilled Murre 1km breeding zone`
+- Layers'/groups' title excludes data provider: Layer should not include provider e.g.
+  "NSIDC". Put this in the dataset citation instead! _EXCEPTION:_ In some cases, we have
+  the same type of data from various providers, e.g.:
+    - `Discipline/Measurement/Provider A/Layer`
+    - `Discipline/Measurement/Provider B/Layer`
 - No fields are populated with "TBD" or "TODO"
 - All layers have style
-- Any known issues with data quality are listed in "Description"
+- Any known issues with data quality are listed in "Description": This should not
+  include spatial mismatch issues, this is covered by our [Disclaimer](/disclaimer.md).
 - Datastore's data is in final state
 - Layers not in unneccessary group
 - Stakeholders approve citation
@@ -110,10 +122,14 @@ A correct QGreenland data pipeline will output data that:
 
 ### Raster layers
 
-- Interpolation algorithm OK? (almost everything is currently bilinear, see notes
-  for exceptions)
-- Grid resolution in layer title or group (incl. units)?
-- When reprojecting (e.g., via `gdalwarp`), ensure that the target resolution is specified (see notes for exceptions)
+- Interpolation algorithm follows standard: Bilinear for continuous values (i.e. real),
+  e.g. temperatures, height. By default, everything is bilinear. Nearest neighbor for
+  discrete (i.e. integer) values, e.g. age of sea ice.
+- Grid resolution in layer title or group (incl. units), e.g. `My layer (1km)`
+- When a layer is reprojected (e.g., via `gdalwarp`), ensure that the
+  target resolution is explicitly specified. Sometimes when the output resolution is not
+  defined, GDAL will choose a resolution that is not appropriate for the data, and this
+  can lead to artifacts.
 
 
 ### Vector layers