This document gives an overview of the data and code included within the resource 'Vegetation-Water Alignment index, Henry's Fork, Idaho case study'. 

DATA ---------------------------------------------------------------------------------------------------
All data used in this analysis is publically available but included in this resource for easy integration with the accompanying code when the whole repository is downloaded. The data folder is organized with multiple subfolders, which indicate the data type stored in each folder. A brief overview of each folder, data source, and processing is provided below. Throughout, missing values are indicated with 'NA'.

hydrology: natural streamflow data for each watershed by WY and in cfs. Summary statistics are provided for each water year.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

melt: melt data in mm for each SNOTEL site used in this study was calculated as the daily reduction in SWE in WY format for water years 1989-2023. Data was accessed using the NRCS soilDB R package (Beaudette et al., 2023). To maintain a consistent number of days for each year, February 29th was removed for leap years, with any new melt added to the March 1 total.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

metrics: metrics included in drought delineation and model selection. An overview of each variable is included below.

- PDSI data is a time series of spatially-averaged GridMet 4km PDSI for the region, including all three watersheds (Abatzoglou, 2011), accessed via https://app.climateengine.org/climateEngine.
	
- MODIS Snow Disappearance Date averaged for each watershed, accessed via https://www.snowcloudmetrics.app/ (Crumley et al., 2020)

- Ag.ET.WY is the time series of the Kimberly-Penman alfalfa reference ET averaged from two US Bureau of Reclamation AgriMet sites located in the valley of the Henrys Fork Watershed (site IDs AHTI and RXGI). Data can be accessed from https://www.usbr.gov/pn/agrimet/wxdata.html. The time series is in WY format for WY 1989-2023 and given in inches.

- AG.1.yr.Moisture.WY is a time series of a short-term drought metric used in this watershed. This metric is the 1-year moving average of the deficit between precipitation across the entirety of the Henrys Fork (averaged from SNOTEL site data) and alfalfa reference ET (Ag.ET.WY) in inches.	

- VWA.CY the calculated VWA for each SNOTEL site and watershed for calendar years 2006-2022.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

ndvi: water year time series of 250m 16-day MODIS NDVI (M*D13Q1) extracted for the pixel of each SNOTEL site location and for the entire watershed. Each time series was created from images filtered using the M*D13Q1 reliability indices to remove low-quality and cloud-covered pixels, and the day of acquisition data was used to assign the correct day for each value. NASA MODIS data was accessed using the MODIStsp R package (Busetto and Ranghetti, 2016). This folder also includes watershed-averaged AVHRR time series for the years 1989-2012 (Vermote et al., 2014). AVHRR data was accessed using Google Earth Engine.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

precip: daily precipitation data (in mm) for each SNOTEL site used in this study in WY format for water years 1989-2023. Watershed average precipitation was calculated as the daily average of all sites associated with a watershed. Data was accessed using the NRCS soilDB R package (Beaudette et al., 2023). To maintain a consistent number of days per year, February 29th was removed for leap years, with any new precipitation added to the March 1 total.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

snow: daily snow depth and SWE data for each SNOTEL site used in this study in WY format for water years 1989-2023 (or the start of station POR) and given in mm. Data was accessed using the NRCS soilDB R package (Beaudette et al., 2023). This data is in standard NRCS convention where the value recorded for a day represents the midnight (00:00) reading at the start of that date. To maintain a consistent number of days per year, February 29th was removed for leap years, with any new snow reflected in the March 1 total.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

temp: bias-corrected daily average temp for each SNOTEL site used in this study in WY format for water years 1989-2023 (or start of station POR). Data was accessed using the NRCS soilDB R package (Beaudette et al., 2023) and corrected as recommended by the NRCS using the sensor history at each site to select the start of the corrected period (Atwood et al., 2023). Temperature data is given in degrees Celsius.



CODE -------------------------------------------------------------------------------------------------------
Each R Markdown file presents a series of analyses relating to the title of the file and the corresponding section of the text. Each file is numbered 1 through 3 in order of appearance and discussion in the text. If the entire resource is downloaded and unzipped, each script can be run independently once the file path is updated to reflect the location of downloaded data. Each Markdown file is commented to outline the analyses, but a brief overview for each file is provided below.

1_TimeTrendAnalysis: This code (1) identifies breakpoints in PDSI; (2) creates pre-drought predictive models for each watershed; (3) tests drought and post-drought model performance; (4) creates Figure 3; (5) performs Mann Kendall trend analysis on climate and vegetation data; (7) creates figures 4.

2_VegetationWaterAlignment: This code (1) creates functions for equations 4, 5, and 6; (2) calculates VWA for each SNOTEL site; (3) calculates watershed VWA using SWI average from all watershed SNOTEL sites and watershed-averaged NDVI; (4) plots figure 5

3_ModelSelectionValidation: This code (1) loads all potential predictors of runoff; (2) ranks post-drought models by AICc; (3) plots best models to create figure 6; (4) performs leave-one-out cross-validation with new and traditional models to assess predictive performance; (5) calculates performance metrics; (6) plots validation results as seen in figure 7.


REFERENCES -----------------------------------------------------------------------------------------------
Abatzoglou J. T. (2011). Development of gridded surface meteorological data for ecological applications and modelling. International Journal of Climatology. doi: 10.1002/joc.3413.

Atwood, J., Domonkos, B., Hill, K., Brosten, T., DeMarco, T., Hultstrand, M., et al. (2023). Evaluation Of Ysi Temperature Correction Equations for Bias-Reducing Snotel Network Temperature Data. 2023 Western Snow Conference

Beaudette, D., Skovlin, J., Roecker, S., Brown, A. (2023). soilDB: Soil Database Interface. R
  package version 2.7.7. 

Busetto, L., Ranghetti, L., (2016) MODIStsp: An R package for automatic preprocessing of MODIS Land Products time series, Computers & Geosciences, Volume 97, Pages 40-48, ISSN 0098-3004, https://doi.org/10.1016/j.cageo.2016.08.020. URL https://github.com/ropensci/MODIStsp/.

Crumley, R. L., Palomaki, R. T., Nolin, A. W., Sproles, E. A., & Mar, E. J. (2020). SnowCloudMetrics: Snow Information for Everyone. Remote Sensing, 12(20), 3341. https://doi.org/10.3390/rs12203341

Vermote, E., Justice, C., Csiszar, I., Eidenshink, J., Myneni, R., Baret, F., Masuoka, E., Wolfe, R., Claverie, M., and NOAA CDR Program (2014): NOAA Climate Data Record (CDR) of Normalized Difference Vegetation Index (NDVI), Version 4. NOAA National Climatic Data Center. doi:10.7289/V5PZ56R6