ET analysis

Interpretting trends in ET from NHM and WRF-Hydro

Results from the national models show significant and consistent negative trends in evapotranspiration throughout much of the southwest US over the time period 1983-2020. The trends are particularly consistent in the Southwest Desert, Central Rockies, and California-Nevada hydrologic regions. These trends appear to be in conflict with some of what is reported in the literature, which shows that the trends in ET are positive across the Upper Colorado River Basin and elsewhere. Although they appear to be in line with trends reported by some source ().

This analysis aims to address some specific questions that will help us interpret trends in ET and how they comport this those reported in the literature. Below are a few questions guiding the analysis:

1. Are trends in ET considerably different between the two models?
2. To what extent are trends calculated for hydrologic regions masking internal heteorogeneity that shows distinct patterns?

---

---

Question 1) Are trends in ET considerably different between the models?

Hydrologic region

We’ll focus analysis on hydrologic regions that have snow dominated HUCs as we suspect that might be an important part of the story, basically removing the SE aggregated region from the analysis.

Looking at ET trends by hydrologic region, there are differences between the models, but there don’t appear to be glaring differences in water year ET in the western US, both models show negative trends in all western regions (PNW, CA-NV, Columbia-Snake, Central Rockies, and SW Desert). In general, the trends from NHM-PRMS are slightly more negative except in the SW Desert.

HUC12

Looking at comparison of models by HUC12: - NHM generally shows higher magnitude changes compared to WRF-Hydro (PNW, Columbia-Snake, Great Lakes, Midwest) - In the Central Rockies, CA-NV, and SW Desert, there is disagreement among HUCs where NHM is showing strong negative trends and WRF-Hydro is showing near zero - In the Northeast and Great Lakes, there is disagreement among HUCs where NHM is showing strong positive trends and WRF-Hydro is showing near zero

Mapping each model’s ET results

Median SWE max

WRF-Hydro ET Slope

## Reading layer `western_vm_reg' from data source 
##   `/Users/ggorski/Library/CloudStorage/OneDrive-DOI/National IWAAs/National Water Supply/1b/analysis/1b-trends-eda/scratch/out/western_vm_reg.shp' 
##   using driver `ESRI Shapefile'
## Simple feature collection with 3 features and 2 fields
## Geometry type: MULTIPOLYGON
## Dimension:     XY
## Bounding box:  xmin: -2339975 ymin: 814187 xmax: -522568.4 ymax: 2412483
## Projected CRS: NAD_1983_Albers

NHM-PRMS ET Slope

In general, WRF-Hydro shows more positive ET trends in the Northern Sierra Nevada, Colorado Front Range, and San Juan Mountains of Colorado. NHM-PRMS shows some spotty positive trends in ET along the Front Range, but generally negative ET trends in the San Juans. Both models show negative trends in Central Arizona and New Mexico.

Model comparison by annual SWE max

• A comparison of ET trends by range of SWE max (columns) for different western regions (rows) shows that at low swe (left), the models have better agreement, but at higher SWE there is less agreement.
• With SWE max 400-800 mm in California-Nevada and the Central Rockies (top two rows, middle column), NHM shows negative slopes while WRF-Hydro shows positive or near zero slopes.
• In the Pacific Northwest and the Columbia-Snake, NHM shows positive slopes where WRF-Hydro shows near zero.