One row per intrinsic-potential (IP) model, species, life stage and
life-history form: where the model was built, how it represents flow and
valley confinement, and whether it was validated against fish. The registry
describes models and holds none of their numbers. A model's thresholds and
curve breakpoints are rows of the literature evidence table
(knw_literature_read()) carrying the same model_id, so the variables a
model uses are the variables of those rows. scripts/0280 checks that link
both ways on model, species, stage and life-history form.
Details
citation_key, page and quote support the flow claim: how the model
sizes streams, or, for flow_metric = none, the variables it does use.
cited_source names the original when that quote is a secondary account of
the model (Sheer et al. 2009 describing Burnett's, for example).
flow_surrogate is the metric a model used where its flow_metric was not
available (modelled width standing in for discharge), and differs from it.
A model with a validation other than none also needs validation_result
and its own validation_page, validation_quote and
validation_quote_source; validation_citation_key names the document that
quote is in when it is not citation_key's. A model with
validation = none carries none of them, since an absence cannot be quoted
(say what was read in note). Likewise flow_source is required unless
flow_metric is none; flow_inputs is required when the flow is
modelled; and a none flow has no source, inputs or surrogate. A text quote is at
least 20 characters. model_id is a lower-case identifier that names one
model, and a model x species x stage x form appears once.
The registry is validated before filtering, so a bad row anywhere in the file is an error whichever species is asked for. A requested species absent from the registry is a warning.
Examples
path <- tempfile(fileext = ".csv")
utils::write.csv(data.frame(
model_id = "burnett2007", model = "CLAMS coastal IP", region = "Oregon Coast Range",
species_code = "CO", life_stage = "rearing", life_history_form = NA,
flow_metric = "mean_annual_flow", flow_source = "modelled",
flow_inputs = "gauge;drainage_area;precipitation", flow_surrogate = NA,
valley_confinement = "valley_width_index", validation = "none", validation_result = NA,
citation_key = "key_a", cited_source = NA, page = "4",
quote = "mean annual flow was estimated from drainage area", quote_source = "text",
validation_citation_key = NA, validation_page = NA, validation_quote = NA,
validation_quote_source = NA, note = NA
), path, row.names = FALSE, na = "")
knw_ip_read(path, species_code = "CO")
#> # A tibble: 1 × 23
#> model_id model region species_code life_stage life_history_form flow_metric
#> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
#> 1 burnett2007 CLAM… Orego… CO rearing NA mean_annua…
#> # ℹ 16 more variables: flow_source <chr>, flow_inputs <chr>,
#> # flow_surrogate <chr>, valley_confinement <chr>, validation <chr>,
#> # validation_result <chr>, citation_key <chr>, cited_source <chr>,
#> # page <chr>, quote <chr>, quote_source <chr>, validation_citation_key <chr>,
#> # validation_page <chr>, validation_quote <chr>,
#> # validation_quote_source <chr>, note <chr>