Merge pull request #248 from nsidc/hazelshapiro-patch-1

Create 002-level-of-detail.md

doc/reference/decision-records/002-level-of-detail.md

@@ -0,0 +1,46 @@
+---
+title: "Priority considerations for how specific vs. general objects (esp. observing systems and intermediate products) should be described in the Benefit Tool"
+date: "2024-01-12"
+author: "Hazel Shapiro"
+status: "Proposed"
+---
+
+## Context
+
+When adding information to the Benefit Tool, there is a persistent question about how detailed each object should be.
+For instance, multiple agencies fund and run automated weather stations in Alaska. Should those be referred to as a group
+and generalized as "automated weather stations" or should each agency or organization be listed separately? Does this
+depend on if there is one place to download all AWS data or not?
+
+Answering these question is complex and each evaluation presents its own set of nuanced decisions, especially around
+level of specificity.
+
+## Decision
+
+These decisions will be made on a case-by-case basis with the 'respondents' having the final decision-making ability.
+US AON staff can advise as needed.
+
+The key factors to consider are:
+* Telling the right story with an individual evaluation. For instance, in the above example,
+  if all AWS stations perform at approximately the same level, it may not be important to list them individually.
+* Re-usability of objects. There is a significant benefit to using the same object in multiple evaluations.
+  This points to using very tangible and specific objects.
+* Readability. At a certain point, including too many objects simply makes the diagram unreadable.
+
+The above list is not ranked.
+See decision record 003 for information specific to reanalysis data sets.
+
+## Consequences
+
+* Decisions will be nuanced and adaptable to each scenario.
+* Decisions will require more time and reflection on the part of the respondents and US AON staff
+  than if there was a single piece of guidance.
+* US AON and it's advisors will likely have to revisit these criteria again in the future.
+* Diagrams may require more nuance to interpret
+
+
+## Consent
+
+* Hazel Shapiro
+* Sandy Starkweather
+* Review by Expert Committee on Methods