Making sense of the crazy outlier year of 2026

Bar chart of Rio Grande runoff at Otowi, NM, January–August total each year, 1896–2026, shown as departure from the long-term average of 846,000 acre-feet. Most years vary within a few hundred thousand acre-feet above or below average. 2026 is the lowest on record at 226,387 acre-feet, about 620,000 acre-feet below average. Two gaps in the gage record, 1906–1909 and 1915–1918, are marked as incomplete rather than shown as zero.

Driest Rio Grande flow on record

Updating my “drought and climate change” lecture each fall is a useful exercise – this year more useful than most. Much like the year itself, my slide deck is setting all time records for new data visualizations needed to make sense of the remarkable 2026.

Update

Great questions from the comments. (I love my commenters.) First, a request to include a smoother to make the trend more legible:

Rio Grande runoff at Otowi, Jan–Aug totals 1896–2026 vs. the long-term average, with a smoothed trend curve. The trend has declined since about 1990, reaching its lowest point in 2026 at 226,387 acre-feet — about 620,000 below the 846,000 AF average.

Now with bonus smoother.

Second, an observation that it looks like it tracks the Pacific Decadal Oscillation, a large scale pattern linked to drier conditions in the southwest. Good eye! A quick run of the numbers based on prior-year fall PDO shows a statistically significant but small (r squared explaining 6 – 9 percent of the variance) relationship.

Worth noting that this is all of the water flowing past Otowi, which is a function of climate and all the many human activities upstream (irrigation reducing flow, San Juan-Chama adding to flow). So it’s not solely a climate signal.

Weirdly, when I redid the analysis for the Del Norte gage, which is above the San Luis Valley diversions and should be a clearer climate-alone signal, the relationship disappears. No correlation at all with PDO, which surprised me.

4 Comments

  1. Hey John, outstanding graphic! I am wondering if you might have another graphic that compares the departure from the mean for the Rio Grande and the Colorado River. How closely correlated are they? Thanks, much.

  2. it would be helpful to draw a best fit curve. Many people don’t have the ability to interpret a graphic like that. They just see a lot of noice.

  3. Nancy – Great suggestion! Added another version. Khalil – Good eye! Though my brief dive into that rabbit hole got weird, see explanation above.

Leave a Reply

Your email address will not be published. Required fields are marked *