jemez_draft.Rmd

---
title: "Validating L-band InSAR snow water equivalanet retrievals: A case study in the Jemez Mountains, New Mexico"
journal: "`r rticles::copernicus_journal_abbreviations(journal_name = 'communication')`"
author:
  - given_name: Jack
    surname: Tarricone
    affiliation: "1"
    email: jtarricone@nevada.unr.edu
    corresponding: true
  - given_name: Ryan
    surname: Webb
    affiliation: "2"
  - given_name: HP
    surname: Marshall
    affiliation: "3"
  - given_name: Anne
    surname: Nolin
    affiliation: "1"
  - given_name: Franz
    surname: Meyer
    affiliation: "4"
# If you have more than one corresponding author, add them manually using the following structure (note the commas):
# Two authors: Daniel Nüst (daniel.nuest@uni-muenster.de) and Josiah Carberry (j.carberry@orcid.org)
# Three authors or more: Daniel Nüst (daniel.nuest@uni-muenster.de), Josiah Carberry (j.carberry@orcid.org), and Markus Konkol (m.konkol@wwu.de)
# If the following line is uncommented, the "corresponding: true" above are ignored
#correspongdingauthors: Daniel Nüst (daniel.nuest@uni-muenster.de) and Josiah Carberry (j.carberry@orcid.org)
# If an author is deceased, please mark the respective author name(s) with a dagger '†' and add a further affiliation; put the decease date in the 'address' field", see 'Nikolaus Copernicus' in template.
# If authors contributed equally, please mark the respective author names with an asterisk '*' and add a further affiliation: 'These authors contributed equally to this work.'", see template.
affiliation:
  - code: 1
    address: Graduate Program of Hydrologic Sciences, University of Nevada, Reno, Reno, NV, USA
  - code: 2
    address: Dept. of Civil & Environmental Engineering & Construction Management, University of Wyoming, Laramie, WY, USA
  - code: 3
    address: Department of Geosciences, Boise State University, Boise, ID, USA
  - code: 4
    address: Geophysical Institute, University of Alaska Fairbanks, Fairbanks, AK, USA
abstract: |
  Snow is a critical water resource for the western US and many regions across the globe. However, our ability to accurately measure and monitor changes in snowpack from satellite remote sensing, specifically its water equivalent, is challenged in mountain regions. To confront these challenges, NASA initiated the SnowEx program, a multi-year effort to address knowledge gaps in snow remote sensing. In the winter of 2020, SnowEx acquired UAVSAR L-band interferometric synthetic aperture radar (InSAR) time series to evaluate its capabilities and limitations for tracking changes in snow water equivalent (SWE). It was tested in a range of snow cover, land cover, and climatic conditions across the western US. A more comprehensive understanding of these limitations will allow the snow community to leverage the upcoming NASA-ISRO (NISAR) mission with its L-band radar. This study analyzed two InSAR pairs from the Jemez River Basin (JRB), NM, between February 12-26, 2020. We developed an end-to-end UAVSAR InSAR processing workflow for snow applications. This open-source approach employs a novel data fusion method that merges optical snow-covered area (SCA) with InSAR data. Combining these two remote sensing data sets allows for robust atmospheric correction and proper delineation of snow-covered pixels. For both pairs, we converted phase change values to SWE change estimations between the two data acquisition dates. We then evaluated radar-derived retrievals using a combination of snow pits, meteorological station data, and ground-penetrating radar (GPR) data. The results of this study show that this InSAR approach is effective for measuring changes in SWE, even in relatively warm snow conditions. Future work will refine the InSAR optical fusion technique and investigate the performance impacts of landscape and snowpack characteristics such as liquid water content, forest cover, elevation, slope, and aspect.
bibliography: sample.bib
running:
  title: R Markdown Template for Copernicus
  author: Nüst et al.
# This section is mandatory even if you declare that no competing interests are present.
competinginterests: |
  The authors declare no competing interests.
# See https://publications.copernicus.org/for_authors/licence_and_copyright.html, normally used for transferring the copyright, if needed. 
# Note: additional copyright statements for affiliated software or data need to be placed in the data availability section. 
copyrightstatement: |
  The author's copyright for this publication is transferred to institution/company. 
### The following commands are for the statements about the availability of data sets and/or software code corresponding to the manuscript.
### It is strongly recommended to make use of these sections in case data sets and/or software code have been part of your research the article is based on.
### Note: unless stated otherwise, software and data affiliated with the manuscript are assumed to be published under the same licence as the article (currently Creative Commons 4.0)
availability:
  #code: |
  #  use this to add a statement when having only software code available
  #data: |
  #  use this to add a statement when having only data sets available
  codedata: |
    use this to add a statement when having data sets and software code available
  sample: |
    use this section when having geoscientific samples available
videosupplement: |
  use this section when having video supplements available
authorcontribution: |
  Daniel wrote the package. Josiah thought about poterry. Markus filled in for a second author.
disclaimer: |
  We like Copernicus.
acknowledgements: |
  Thanks to the rticles contributors!
appendix: |
  \section{Figures and tables in appendices}
  Regarding figures and tables in appendices, the following two options are possible depending on your general handling of figures and tables in the manuscript environment:
  \subsection{Option 1}
  If you sorted all figures and tables into the sections of the text, please also sort the appendix figures and appendix tables into the respective appendix sections.
  They will be correctly named automatically.
  \subsection{Option 2}
  If you put all figures after the reference list, please insert appendix tables and figures after the normal tables and figures.
  
  To rename them correctly to A1, A2, etc., please add the following commands in front of them:
  `\appendixfigures` needs to be added in front of appendix figures
  `\appendixtables` needs to be added in front of appendix tables
  
  Please add `\clearpage` between each table and/or figure. Further guidelines on figures and tables can be found below.
output:
  rticles::copernicus_article: 
    highlight: NULL
    keep_tex: true
  bookdown::pdf_book:
    base_format: rticles::copernicus_article # for using bookdown features like \@ref()
---
# Introduction
    In the western US, seasonal mountain snowpack is the primary water source, providing water for about 60 million people (Stewart et al., 2004). The snowmelt flows directly to users downstream, stored in reservoirs to generate hydropower, or is diverted far beyond the catchment boundaries. This makes snowmelt a highly variable but essential resource for the entire western US population, not just the locations near mountains. The snowpack acts as a reservoir in the winter as it accumulates and stores the frozen snowfall (Bales et al., 2006). Every spring, it releases the water as snowmelt, which flows down the catchment to recharge groundwater and fill up man-made reservoirs. Humans use this water for agriculture, industry, ecosystem services, and almost every other facet of modern life. To correctly manage these resources, society needs accurate knowledge of how much water is held in the seasonal snowpack and when that water will be released as snowmelt. \par
  Currently, our estimation of the spatial and temporal variations in mountain SWE remains imprecise (Dozier, 2011). The Natural Resource Conservation Services (NRCS) Snow Telemetry (SNOTEL) network uses snow pillows and courses to monitor snow at point-based locations. This information is then used by National Weather Service (NWS) River Forecast Centers to inform statistically-based streamflow forecasting models, without assimilating any of the remote sensing data products (Pagano et al., 2004). \par
Climate warming is affecting the stationarity of various components of the hydrologic cycle across the globe (Milly et al., 2008). In the dry mountain western US, changes from snowfall to rainfall and less precipitation could significantly impact society by changing the magnitude and timing of the spring streamflow (Harpold et al., 2017a). This, in turn, decreases our ability to predict and forecast streamflow because it violates the statistical assumption of stationarity in the regression-based models (A. A. Harpold et al., 2017; Miller et al., 2011). Therefore, we need to look to remote sensing tools for a more accurate understanding of the spatiotemporal distribution of our mountain snowpack. \par
  There is a rich history of remote sensing of various snowpack properties, dating back to the first Earth observing satellites (Barnes & Bowley, 1968). These properties include snow-covered area (SCA), snow albedo, snow grain size, snowpack wetness, snow depth, and SWE. While the ability to measure these properties provides useful information in the different SWE estimation techniques, there is still no single remote sensing technique that can continually measure SWE from space for mountain snowpack applications (Lettenmaier et al., 2015). \par
Optical remote sensing satellites such as the Moderate Resolution Imaging Spectrometer (MODIS) and Landsat can produce both SCA and fractional snow-covered area (fSCA) maps through a spectral unmixing analysis (Bair et al., 2019; Dozier, 1989; Painter et al., 2009; Rittger et al., 2013). Suborbital lidar (Painter et al., 2016) combined with hyperspectral imaging (A. Nolin et al., 1993; A. W. Nolin & Dozier, 1993) are mature techniques for sensing both depth and fSCA at the watershed scale. To convert these optical measurements into SWE, both require spatially distributed snowpack energy balance models. Like all optical techniques, they are limited by cloud cover, which is frequent in mountain environments. \par
  Passive microwave emissions have been used to estimate SWE (Dong et al., 2005; Foster et al., 2005; Vuyovich et al., 2014). These measurements have a large spatial resolution (25x25 km), and cannot capture the topographic heterogeneity of mountain environments. This technique is also limited by deep snowpack (>1m), and the presence of liquid water within the snowpack, which are common in mountain environments. While passive microwave isn’t suited for mountain environments, other radar techniques should be considered. \par
  Synthetic Aperture Radar (SAR) is an active remote sensing technique that shows promise for snowpack monitoring. It addresses the two main deficiencies in both optical and passive microwave; it can penetrate through clouds and has a spatial resolution on the scale of tens of meters instead of kilometers. Recently, there has been substantial growth in the use of SAR data for many Earth science disciplines due to the increase in available sensors, ease of data processing and image focusing, and access to cloud-based data products. The data volume will only increase with the planned launch of NISAR, Surface Water Ocean Topography Mission (SWOT), Radar Observing System for Europe L-Band  (ROSE-L), and the continuation of Sentinel-1 mission. \par
  SAR backscatter has been used to estimate SWE at shorter wavelengths (Ku, Ka, X) (Rott et al., 2011; J. Shi & Dozier, 2000; Jiancheng Shi et al., 1994; Tsang et al., in review). This method requires a complex radiative transfer model with input parameters that include snow density, snow temperature, and grain size, which are challenging to estimate over large spatial scales precisely. SAR is robust for measuring snow wetness (Nagler, 1996; Nagler et al., 2016), and new backscatter methods are being developed to measure snow depth at C-band (Lievens et al., 2019, 2021) \par
  Recently, the use of InSAR to estimate SWE has become an area of interest because of the higher temporal (12 days) frequency and L-band wavelength of NISAR. InSAR uses the differences in radar phase between subsequent overpasses to estimate surface displacement. The SWE InSAR theory, initially proposed by Guneriussen et al. (2000), relates changes in the interferometric phase of a radar signal to changes in dry snow on the ground between acquisitions. This technique is limited to dry snow because the presence of liquid water vastly increases the dielectric permittivity value of snow and thus reduces the backscattering coefficient. These dielectric losses cause the signal to attenuate in the snowpack and not scatter off the snow-ground interface, therefore breaking down the physical basis of the technique (Nagler, 1996). \par
  After publication of the initial theory, follow-up studies were performed by Rott et al. (2003) in Austria and Deeb et al. (2011) on Alaska’s north slope, both using ERS-1 C-band radar. Leinss et al. (2015) conducted an intensive season-long ground-based dual-frequency (X and Ku)   interferometric experiment in Finland. They found the method was successful for continually measuring SWE in dry taiga snow, but that liquid water and vegetation would quickly cause loss of coherence. The Sentinel-1 C-band radar was used more recently, leveraging the more consistent overpass repeat cycle (Conde et al., 2019). These orbital InSAR studies showed promise for estimating SWE but lacked sufficient temporal length and variety of vegetation, topography, and snowpack characteristics to thoroughly understand the technique’s limitations and synergies with other types of snow measurements. \par

# Methods
## Study Area Description

Located in northern New Mexico, the Jemez Mountains, and Jemez River are on the southern extent of the Rocky Mountains. The Jemez River research site is the furthest south of the 13 SnowEx 2020 campaign sites (Figure 2) and has a varied and unique hydroclimatic, topographic, and vegetated environment. Within the Jemez River Basin lies Valles Caldera, a 25-km wide volcanic structure dating back about 1.2 million years. Within Valles Caldera, the Valle Grande is an extensive open grassland where field measurements took place for this study. Many resurgent lava domes form peaks over the grassy valleys, the highest of Redondo Peak (11,253 ft).

```{r echo=FALSE, out.width='100%', echo = FALSE, fig.pos="H", fig.cap = "(a) Map showing the location of the UAVSAR acquisition (black outline) in the Jemez Mountains, NM. (b) DEM of the UAVSAR acquisition provided by JPL, with a black rectangle outlining Valle Grande and the surrounding hill slopes. (c) DEM of Valle Grande with snowpit locations shown by black triangles, meterologic stations shown by blue circles, and the GPR transect shown as a black line. (d) A close up of the GPR transect with the HQ Met snow pit and meterologic station displayed."}
knitr::include_graphics("jemez_map_new.png")
```





## Data Description

For this study, we are analyzing two UAVSAR L-band InSAR image pairs in the Jemez River Basin, NM from 2/12 - 2/19 and 2/19 - 2/26. UAVSAR is full polarimetric L-band radar deployed on a NASA Gulf Stream III aircraft, traditionally flown at 45,000 ft with a 16 km nominal range swath width (Rosen et al., 2006). The flight management system and  GPS are precise enough for conducting InSAR measurements. On these dates, field teams measured snowpack density, depth, wetness, temperature, stratigraphy, and grain size. Researchers also collected data using a ground-penetrating radar (GPR), which will be used to validate the InSAR-derived changes in SWE at scales from meters to a kilometer. Landsat fSCA data from 2/18/2020 will be used to identify snow-covered pixels. Within the radar swath, there are two SNOTEL sites (Garita Peak & Quemazon), and three meteorologic stations, which will be used to help validate the radar SWE retrievals. 

```{r echo=FALSE, out.width='100%', echo = FALSE, fig.pos="H", fig.cap = "(a) Figure showing InSAR data from ....... "}
knitr::include_graphics("cor_unw_amp_v1.png")
```

## Atmospheric Correction

## New Incidence Angle Data

The following settings can or must be configured in the header of this file and are bespoke for Copernicus manuscripts:

- The `journal` you are submitting to using the official abbreviation. You can use the function rticles::copernicus_journal_abbreviations(name = '...')` to search the existing journals.

- Specific sections of the manuscript:

  - `running` with `title` and `author`

  - `competinginterests`

  - `copyrightstatement` (optional)

  - `availability` (strongly recommended if any used), one of `code`, `data`, or `codedata`

  - `authorcontribution`

  - `disclaimer`

  - `acknowledgements`

See the defaults and examples from the skeleton and the official Copernicus documentation for details.

**Please note:** Per [their
guidelines](https://publications.copernicus.org/for_authors/manuscript_preparation.html),
Copernicus does not support additional \LaTeX{} packages or new \LaTeX{}
commands than those defined in their `.cls` file. This means that you cannot add any extra dependencies and a warning will be thrown if so.  
This extends to syntax highlighting of source code. Therefore this template sets
the parameter `highlight` in the YAML header to `NULL` to deactivate Pandoc
syntax highlighter. This prevent addition of external packages for highlighting
inserted by Pandoc. However, it might be desirable to have syntax highlight
available in preprints or for others reasons. Please see
`?rmarkdown::pdf_document` for available options to activate highlighting.

**Important**: Always double-check with the official manuscript preparation
guidelines at
[https://publications.copernicus.org/for_authors/manuscript_preparation.html](https://publications.copernicus.org/for_authors/manuscript_preparation.html),
especially the sections "Technical instructions for LaTeX" and "Manuscript
composition". Please contact Daniel Nüst, `daniel.nuest@uni-muenster.de`, with
any problems.

# Content section one

## Subsection Heading Here

Subsection text here. 

### Subsubsection Heading Here

Subsubsection text here. 

# Content section with citations

See the [R Markdown docs for bibliographies and citations](http://rmarkdown.rstudio.com/authoring_bibliographies_and_citations.html).

Copernicus supports biblatex and a sample bibliography is in file `sample.bib`.
Read [@Feynman1963118], and [see @Dirac1953888].

# Content section with R code chunks

```{r, echo = FALSE}
sum <- 1 + 41
```

You should always use `echo = FALSE` on R Markdown code blocks as they add formatting and styling not desired by Copernicus.
The hidden workflow results in `r sum`.

You can add verbatim code snippets without extra styles by using ` ``` ` without additional instructions.

```
sum <- 1 + 41
```

# Content section with list

If you want to insert a list, you must

- leave

- empty lines

- between each list item

because the `\tightlist` format used by R Markdown is not supported in the Copernicus template. Example:

```
- leave

- empty lines

- between each list item
```

# Examples from the official template

## FIGURES

When figures and tables are placed at the end of the MS (article in one-column style), please add \clearpage between bibliography and first table and/or figure as well as between each table and/or figure.

### ONE-COLUMN FIGURES

Include a 12cm width figure of Nikolaus Copernicus from [Wikipedia](https://en.wikipedia.org/wiki/File:Nikolaus_Kopernikus.jpg) with caption using R Markdown.

```{r, out.width = "8.3cm", echo = FALSE, fig.cap = "one column figure"}
knitr::include_graphics("Nikolaus_Kopernikus.jpg")
```

### TWO-COLUMN FIGURES

You can also include a larger figure.

```{r, out.width = "12cm", echo = FALSE, fig.cap = "two column figure"}
knitr::include_graphics("Nikolaus_Kopernikus.jpg")
```

## TABLES

You can ad \LaTeX table in an R Markdown document to meet the template requirements.

### ONE-COLUMN TABLE

\begin{table}[t]
\caption{TEXT}
\begin{tabular}{l c r}
\tophline

a & b & c \\
\middlehline
1 & 2 & 3 \\

\bottomhline
\end{tabular}
\belowtable{Table Footnotes}
\end{table}

### TWO-COLUMN TABLE

\begin{table*}[t]
\caption{TEXT}
\begin{tabular}{l c r}
\tophline

a & b & c \\
\middlehline
1 & 2 & 3 \\

\bottomhline
\end{tabular}
\belowtable{Table footnotes}
\end{table*}

## MATHEMATICAL EXPRESSIONS

All papers typeset by Copernicus Publications follow the math typesetting regulations given by the IUPAC Green Book (IUPAC: Quantities, Units and Symbols in Physical Chemistry, 2nd Edn., Blackwell Science, available at: http://old.iupac.org/publications/books/gbook/green_book_2ed.pdf, 1993).

Physical quantities/variables are typeset in italic font (t for time, T for Temperature)

Indices which are not defined are typeset in italic font (x, y, z, a, b, c)

Items/objects which are defined are typeset in roman font (Car A, Car B)

Descriptions/specifications which are defined by itself are typeset in roman font (abs, rel, ref, tot, net, ice)

Abbreviations from 2 letters are typeset in roman font (RH, LAI)

Vectors are identified in bold italic font using \vec{x}

Matrices are identified in bold roman font

Multiplication signs are typeset using the LaTeX commands `\times` (for vector products, grids, and exponential notations) or `\cdot`

The character * should not be applied as multiplication sign

## EQUATIONS

### Single-row equation

Unnumbered equations (i.e. using `$$` and getting inline preview in RStudio) are not supported by Copernicus.

\begin{equation}
1 \times 1 \cdot 1 = 42
\end{equation}

\begin{equation}
A = \pi r^2
\end{equation}

\begin{equation}
x=\frac{2b\pm\sqrt{b^{2}-4ac}}{2c}.  
\end{equation}

### Multiline equation

\begin{align}
& 3 + 5 = 8\\
& 3 + 5 = 8\\
& 3 + 5 = 8
\end{align}

## MATRICES

$$
\begin{matrix}
x & y & z\\
x & y & z\\
x & y & z\\
\end{matrix}
$$

## ALGORITHM/PROGRAMMING CODE

If you want to use algorithms, you need to make sure yourself that the \LaTeX packages `algorithms` and `algorithmicx` are installed so that `algorithm.sty` respectively `algorithmic.sty` can be loaded by the Copernicus template. Both need to be available through your preferred \LaTeX{} distribution. With TinyTeX (or TeX Live), you can do so by running `tinytex::tlmgr_install(c("algorithms", "algorithmicx"))`

```{r, echo = FALSE, eval = tinytex::is_tinytex()}
tinytex::tlmgr_install(c("algorithms", "algorithmicx"))
```

Copernicus staff will no accept any additional packages from your LaTeX source code, so please stick to these two acceptable packages. They are needed to use the example below

\begin{algorithm}
\caption{Algorithm Caption}
\label{a1}
\begin{algorithmic}
\STATE $i\gets 10$
\IF {$i\geq 5$} 
        \STATE $i\gets i-1$
\ELSE
        \IF {$i\leq 3$}
                \STATE $i\gets i+2$
        \ENDIF
\ENDIF
\end{algorithmic}
\end{algorithm}

## CHEMICAL FORMULAS AND REACTIONS

For formulas embedded in the text, please use `\chem{}`, e.g. \chem{A \rightarrow B}.

The reaction environment creates labels including the letter R, i.e. (R1), (R2), etc.

- `\rightarrow` should be used for normal (one-way) chemical reactions

- `\rightleftharpoons` should be used for equilibria

- `\leftrightarrow` should be used for resonance structures

\begin{reaction}
A \rightarrow B \\
\end{reaction}
\begin{reaction}
Coper \rightleftharpoons nicus \\
\end{reaction}
\begin{reaction}
Publi \leftrightarrow cations
\end{reaction}

## PHYSICAL UNITS

Please use `\unit{}` (allows to save the math/`$` environment) and apply the exponential notation, for example \( 3.14\,\unit{km\,h^{-1}} \) (using LaTeX mode: `\( 3.14\,\unit{...} \)`) or \unit{0.872\,m\,s^{-1}} (using only `\unit{0.872\,m\,s^{-1}}`).

\conclusions[Conclusions]

The conclusion goes here.
You can modify the section name with  `\conclusions[modified heading if necessary]`.