### The data presented here were developed using the methods outlined in Blythe and Schmidt 2018 (see citation below) ###

Citation: Blythe, T. L., & Schmidt, J. C. (2018).
Estimating the natural flow regime of
rivers with long-standing
development: The Northern branch of
the Rio Grande. Water Resources
Research, 54. https://doi.org/10.1002/
2017WR021919

Data are presented as streamflow values at a location in cubic-feet per second. All data from 1900 to 2010 have been 
reviewed and are considered final estimates of the natural streamflow. All data after 2010 are considered provisional and
need review. NA values exist in the final estimates for two reasons 1) there were no initial input data at the most upstream
gage for the model, and 2) the model estimated negative streamflow. Natural flows for the three tributaries with two or more 
gages are also included. The two comparison datasets are from the US Department of Agriculture and the US Army Corps of 
Engineers (see descriptions and limitations below).

Included in this dataset:

"Rio_GrandeNatural_Final_Mass_Balance.csv"
- This file contains the final output for the primary mainstem gages. Column labels include the gage names for location. These
  values could be considered a conservative maximum estimate.

"Rio_GrandeNatural_LowerErrorBand.csv"
- This file contains the lower limit of the estimated error based on average un-gaged area. These values are presented as
  daily values because they are estimated using the same modeling procedure but using less gages (i.e. greater un-gaged areas).
  Data structure is the same as "Rio_GrandeNatural_Final_Mass_Balance.csv"

"Rio_GrandeNatural_UpperErrorBand_CalendarYr.csv"
- This file is the upper limit of the estimated error based on average un-gaged area. These values are given as annual values for
  the calendar year. These are not presented as daily values because they are based on a regression between estimated natural and
  observed streamflow (daily values have significant autocorrelation and did not provide a good least-squares fit). We used a
  regression approach because the current number of gages is the maximum number of gages that has existed on the Rio Grande. Our
  uncertainty analysis provides less uncertainty with more gages, however, it is impossible to determine how uncertainty would
  be affected if we added an infinite number of gages to the mainstem. This is our best method for estimating the upper uncertainty
  limit as the number of gages reaches infinity - or, the un-gaged area approaches zero. Data structure is the same as 
  "Rio_GrandeNatural_Final_Mass_Balance.csv"

"Rio_GrandeNatural_UpperErrorBand_WaterYr.csv"
- Same as "Rio_GrandeNatural_UpperErrorBand_CalendarYr.csv" except values are calculated for water year rather than calendar year.

"Conejos_RiverNatural_FinalMassBalance.csv"
- This file is the model results for the estimated natural flow of the Conejos River. Zero values occur when there were insufficient 
  upstream input data. When used with the mainstem values, a zero value is treated as an un-gaged tributary.

"Rio_ChamaNatural_FinalMassBalance.csv"
- This file is the model results for the estimated natural flow of the Rio Chama. Zero values occur when there were insufficient 
  upstream input data. When used with the mainstem values, a zero value is treated as an un-gaged tributary.

"Jemez_RiverNatural_FinalMassBalance.csv"
- This file is the model results for the estimated natural flow of the Jemez River. Zero values occur when there were insufficient 
  upstream input data. When used with the mainstem values, a zero value is treated as an un-gaged tributary.

### Comparison Datasets ###

These data were provided by federal agencies as estimates of Rio Grande Natural Flows.

"USACE_URGWOM_FIN_1975-2010.csv"
- This file is provided by the Upper Rio Grande Water Operations Model (URGWOM) developed by the US Army Corps of Engineers,
  Albuquerque District. Columns include abreviated gage descriptions which are as follows:
  "CntrlGgInflw" = Rio Grande at Albuquerque, NM
  "ChmtGgInflw" = Rio Chama streamflow at the mouth of the Rio Chama
  "DlNrtGgInflw" = Rio Grande near Del Norte, CO
  "EmbdGgInflw" = Rio Grande near Embudo, NM
  "LbtsGgInflw" = Rio Grande near Lobatos, CO
  "OtwGgInflw" = Rio Grande at Otowi Bridge, NM
  "RGAtAnthnyGgInflw" = Rio Grande near El Paso, TX
  "SnMrclFldwyGgInflw" = Rio Grande at San Marcial Floodway, NM
  These data are provided via a reconfigured URGWOM to route flow through reservoirs (Run-of-River), and
  remove all diversions. The estimates do not include natural un-gaged inflow below Otowi Bridge nor natural transmission losses.
  These data are considered provisional and are not regularily produced by the URGWOM. For further questions or concerns regarding
  the production, use, and distribution of this data, please contact Marc Sidlow - P.E. U.S. Army Corps of Engineers, Albuquerque
  District.

"USDA_NRCS_AdjustedFlows.csv"
- This file contains estimates of flow adjusted for human uses by the Natural Resource Conservation Service. These data are publicly
  available online via the USDA report generator. Columns contain gage names and numbers for location. These data are not accompanied
  by any rigorous explanation of methods or metadata. Futher questions regarding these data should be directed to the U.S. Department
  of Agriculture.