Trends in 20th century drought over the U.S. |
In my previous post I illustrated how decadal climate variability over the last millennium affected populations of pinyon pine in the US south-west. I will now narrow the focus to the 20th century and ask whether droughts are better of worse. (See this earlier post on the intricacies of defining a drought and how to measure one.) For the answer I’ll turn to a study published in Geophysical Research Letters earlier this year by two researchers at the University of Washington.
The researchers used a hydrological model to simulate soil moisture conditions and runoff over the continental U.S. for the period 1925–2003. Good enough records of soil moisture just don't exist over U.S., not even today, so a model is used to reconstruct the record (recall that the best estimate of an agricultural drought is obtained through this type of numerical model, rather than an index based on the climate). They used the Variable Infiltration Capacity model, which calculates the energy and water balance at each point based on climatic inputs and a description of the land surface (soil, vegetation, topography). They examined agricultural and hydrological droughts, using estimates of soil moisture and runoff, spat out by the model, as drought indicators. They then used a statistical test (the Mann-Kendall test, if you really want to know) to see whether there were any trends in the model output. Indeed there were.
Over the 20th century, the soil became wetter over the majority of the country, while drying predominantly occurred in the Southwest. Trends in annual runoff were similar. Decreases in drought severity occurred in the Northeast, Great Lakes, Lower Mississippi, and the Pacific Northwest; increases in severity were located mostly in Texas, the Southwest and intermountain West. The eastern U.S. and Midwest saw a reduction in drought frequency, while the Southwest saw a little increase. Drought duration decreased in the Great Lakes, Mississippi basin, and Pacific Northwest, while duration increased in parts of the west including California and the interior.
While there has been a general increase in precipitation over the U.S., some areas still exhibited worse droughts. In those parts of the country that showed a decrease in soil moisture over the century (locations in the Southwest and the West), the increase in precipitation was offset by higher temperatures, causing more water to be evaporated.
The study made no mention of climate change, but it does illustrate the importance of making regional assessments of any climate change – the hydrology of different regions is different, and so the response to any changes in climate may differ too. The study also demonstrates that we can’t take a simple approach to predictions of drought, like saying more rain means less drought. The whole water cycle must be taken into account.
Labels: water resources

2 Comments:
John Fleck says -
One of the important messages to me in the Andreadis paper is that the area of the U.S. that is most water-stressed is the area that saw an increase in drought. But maybe that's my own geographical bias showing. :-)
I'd also be curious about your thoughts on how this result compares to Aiguo Dai's findings: Journal of Hydrometeorology, Volume 5, Issue 6 (December 2004), pp. 1117–1130
That JHydromet paper is hefty, so I only want to skim it. Here's my quick take.
They say the U.S. got wetter and less droughty over the 20th century. This included the Southwest, which is opposite from the conclusion I reported above.
So, why the difference you ask?
The JHydromet paper used Palmer, which they say is for meterological drought right off the bat. They do go on to say what environmental factors go into the index, and burried in the body they go on to say what it doesn't do so well. The big one in my mind is evapotranspiration, with both wind and vegetation lacking proper treatment.
The VIC model I'm sure does ET much better. And this is very important. Consider the early 2000s drought with the 1950s one - the latter one was worse because of temp (ie. because of ET).
Also, the study I write about is carefully talking about hydro and ag droughts.
Maybe my bias as an engineer and hydrologist is showing (the GRL paper is by two engineers; and I make hydrological models like VIC). I put more trust in the GRL paper. It has a stronger biophysical basis than they gray/black box of Palmer.
Post a Comment
Subscribe to Post Comments [Atom]
Links to this post:
Create a Link
<< Home