Utah Is Not the Second-Driest State: A Lesson in Questioning Persistent Assumptions about Hydrology

Ever since I entered the water profession and began practicing in Utah about 15 years ago, I have repeatedly heard that “Utah is the second-driest state in the country.” The statement is arresting and plausible given how little precipitation we get, and it has been long propagated by state agencies, water managers, researchers, and the media to advocate prudent water management among the state’s 3.3 million residents.

Although I agree with the intent of the statement, I take issue with it for two reasons. The first is minor: as I will show, Utah is actually the third-driest state in terms of precipitation (which still makes it very dry and very much in need of prudent water management). The second is far more important: why has a readily verifiable hydrologic fact persisted in error, especially among its own stakeholders, and what else that we think we know about hydrology is wrong?

Despite diligent research into published material and conversations with state water experts, I am not sure where the second-driest phenomenon started or why Utahns have latched on to it so firmly. I would almost call it suspicious, save for all the good it has done in raising awareness about our precious water resources, especially against the backdrop of an exceptional regional drought (Wheeler et al. 2022; Abbott et al. 2023) and the fastest urban growth in the country (Associated Press 2021). Table 1 presents a sample of such language from various groups, appearing in everything from websites to master’s theses to the title of one community’s water conservation plan. These organizations have done great work for Utah’s water community, so my purpose is not to criticize them but to demonstrate just how common the phrase has become as a starting point for water research, management, and policy.

Given the consistency with which I and other Utah colleagues have encountered “second driest” in our careers, I figure that the saying has been circulating for decades. The statements never cite any source or data set or method; apparently Utah’s second-place position is so obvious it needs no citation. If it had any basis to begin with, it has been lost and is now nothing more than an assumption, and a long-held one at that. Perhaps it is one of those things that, if repeated enough, people will eventually believe.

A couple personal examples will illustrate how inveterate the phrase is. In 2019, I was part of a team consulting the Utah Division of Water Resources (2019) on new water conservation goals. Discussion about Utah’s position as the second-driest state frequently came up, but try as we might, we could not find any supporting data. As a project team, we agreed to stop saying it and to strike any such language from our report. In 2020, the phrase surfaced in a proposed bill amending the state water policy and, had it not been for a comment I submitted about it during the legislative session, I think the erroneous ranking would have been immortalized in statute. It now reads “Utah is one of the most arid states,” and the correction appears in Line 36 of the official record (Utah State Legislature 2020).

Now, to settle the matter of the driest states (at least for the time being, until the changing climate says otherwise), let us consider the data.

As the comments in Table 1 suggest, rankings about driest states might be based on long-term average precipitation within the state boundaries, which is easy to both understand and verify. Short-term measures like annual precipitation do not adequately describe long-term climate, per-capita water supply is not consistently reported among states, streamflow is a tangible but data-intensive indicator, and more complex measures like the Palmer Drought Severity Index are not readily grasped by the public. Precipitation, like groundwater and streams, does not respect political boundaries, and a purist might argue that a watershed-based characterization of dryness, such as streamflow trends, is more accurate than a state-by-state comparison. So for communicating climate to the public, which is the intention driving the statements in Table 1 in the first place, precipitation within state lines will have to do.

The National Oceanic and Atmospheric Administration (NOAA) serves up statewide precipitation averages in its Climate at a Glance tool (NOAA 2022). The data originate in the US Climate Divisional Database (NClimDiv) (Vose et al. 2014a, b), which consists of observations taken from individual weather stations, interpolated and bias-corrected into a 5-km grid, and aggregated to various regions, including states. In climatology, the standard is 30-year normals, computed in decadal installments (Arguez and Vose 2011; Arguez et al. 2012).

From this data set, 30-year average precipitation for the last five periods consistently shows the driest states to be Nevada (NV), Arizona (AZ), Utah (UT), New Mexico (NM), and Wyoming (WY), in that order [Fig. 1(a)]. Gridded 30-year precipitation normals from Oregon State University’s PRISM Climate Group (2021) confirm the rankings for 1991–2020. By this measure, Utah is the third-driest state and has been for a while.

Using the same NOAA data and expanding the analysis to consider annual precipitation in all 50 states in each of the 30 years from 1991 to 2020 offers some nuance for the recent rankings. Table 2 presents data from the five states in question. The driest state was most frequently Nevada (21 years); second driest was Arizona (17 years); third driest was Utah (10 years); fourth driest was also Utah (14 years); and the fifth driest was Wyoming (16 years). Stated differently, Utah’s most common rank was fourth driest (14 years), then third driest (10 years), then second and fifth driest (3 years each). The average rankings during this time were Nevada, 1.4; Arizona, 2.3; Utah, 3.6; New Mexico, 3.9; and Wyoming, 5.1. By this analysis, Utah is again third (or fourth!), but rarely second.

To be fair, I did find one alternative data set suggesting Utah is the second-driest state. It was posted by the Western Regional Climate Center (n.d.) and refers to the Historical Climatology Series, a now apparently retired NOAA program whose original precipitation report I could not locate. Without being able to verify its origins, I stand by the data in Fig. 1(b).

Hydrology always has some uncertainty. Examining Fig. 1(b), I concede that in some periods the second and third places are close. Still, given the data, I think we may comfortably conclude that Utah is not the second-driest state after all. I suppose it could be the second-driest state by some alternative precipitation data set or by some other long-term metric besides average precipitation that everyone has meant all along. If so, then by all means, show me the source—and then please cite it when you make your claim to others.

Regardless of Utah’s exact rank among the driest US states, my point is that an erroneous hydrologic “fact” seems for decades to have been perpetuated—and widely accepted—by its own stakeholders, with little basis and few questions. Whether Utah is second or third driest makes little difference in water messaging or water management, so in this case the consequences are probably benign.

But the thought should make us pause. What else that we think we know about hydrology is actually wrong? What facts are we not verifying because we have used them for so long? What are we assuming without questioning and leading the public to believe? What unintended consequences are we overlooking (Sowby and Hotchkiss 2022)?

In the interest of advancing the field, we should occasionally reexamine “the many hypotheses that have become standard and have not been sufficiently challenged” (Di Baldassarre et al. 2016). Questioning assumptions and seeking authoritative evidence are still fundamental scientific practices for any research or engineering activity, even when the facts seem obvious. Conditions change, new information arises, and people make mistakes, so we must always be seeking new knowledge.

All data, models, or code generated or used during the study are available from the corresponding author by request.