1/ New study by Ben Cook (@DustyBowl), Jason Smerdon, and me in @NatureClimate. 2000-21 drought was likely the driest 22 years in >1200 years in SW North America. CMIP6 climate models suggest ~40% of the severity was due to human-caused climate trends. nature.com/articles/s4155… 
2/ The cause of drier-than-average soils in the 2000s was low precipitation, but high temperature and vapor-pressure deficit (combined effect of temperature and humidity on atmospheric aridity) made the 2000s drought more severe.
3/ Summer soils were drier than average in 18 of the past 22 years in SWNA. After wet years in 2017 and 2019 it seemed the early 2000s drought might be petering out, but dry conditions reappeared in 2020 and 2021, making clear that the 2000s drought is still in full swing.
4/ Severely dry soil in summer 2021 was caused by dry and hot conditions from summer 2020 through summer 2021. The 2021 monsoon was wet in the SW US and N Mexico, but summer 2021 was exceptionally hot and arid across most of the West.
5/ The uniquely dry nature of 2020-21 can be seen in the flow out of the Upper Colorado River Basin. Estimated Naturalized flow in water-year (WY) 2021 is 4th lowest since at least 1906 and WY 2020-21 saw the lowest 2-yr mean on record. usbr.gov/lc/region/g400…
6/ The extreme summer aridity in 2020 and 2021 is also exemplified in the annual record of western US forest area burned.
7/ In summer-fall 2021, the US Drought Monitor classified a greater proportion of the western US as under extreme or exceptional drought than at any other time since records began in 2000.
8/ For a longer-term perspective on the 2000s drought we use tree-ring records. Scientists have collected tree-ring records in western N America for a century and many have generously made their data public. In many places tree-ring widths correspond to soil moisture.
9/ We use tree-ring records from thousands of trees at hundreds of sites to reconstruct summer soil moisture. The reconstructions are accurate during their period of overlap with climate observations, giving confidence in their fidelity during centuries prior to climate records.
10/ Our reconstruction shows that regionally averaged summer soil moisture across the SW North American study region over 2000-2021 was probably drier than any other 22-year period since at least 800 AD. 2021 ranks as 12th driest during this time (2002 was 11th).
11/ Comparing the progression of the 2000s drought to that of the big megadroughts from last millennium, it is clear that the current event has been on a megadrought trajectory since its inception.
12/ Just like the 2000s, all of the reconstructed megadroughts had occasional wet years. In 2000-2021, 18 years were drier than average. Only two megadroughts contained 22-year periods with more dry years than 2000-2021.
13/ Most of Southwestern North America did not *locally* experience 2000-2021 as the driest 22-yr period in 1200 years. That result is specific to the regional average. But 2000-2021 ranked among the top 5 local 22-yr drought periods across most of the region.
14/ In fact, a greater proportion of the study region experienced top-5 22-yr drought severity in 2000-2021 than during any other period in the reconstruction.
15/ How much of the current drought’s severity has been due to human-caused climate trends? We assess human-caused climate trends as the average trends as simulated by 29 climate models used in the recent CMIP6.
16/ To assess the contribution of human caused climate trends to recent soil-moisture anomalies, we first consider a hypothetical world in which these trends never occurred.
17/ We then modeled the soil-moisture response to this hypothetical no-human climate. The difference between the hypothetical soil moisture and that modeled from observed climate is from human-caused trends. In 2021, human-caused trends made up 19% of the soil-moisture anomaly.
18/ Natural climate variability dominates interannual variability but tends to average out over longer periods of time. Over 2000-2021, 42% of the drought’s severity is accounted for by human-caused climate trends as represented by the CMIP6 29-model mean.
19/ The exact magnitude of the human-caused contribution is highly uncertain, largely due to model disagreement on precipitation trends, but all 29 CMIP6 models suggest a net human-caused drying effect as of 2021 due to the combined effects of temperature, humidity, & precip.
20/ Had the sequence of wet-dry years occurred as observed but without the human-caused drying trend, we estimate that the 2000s would have still been dry, but not on the same level as the worst last millennium’s megadroughts.
21/ In fact, without human-caused climate trends, 2000–2021 would probably not even classify as a single drought event due to what would have been wet conditions in 2004-2005.
22/ Will the 2000s drought reach the 23-year duration of the 1500s megadrought? Replacing 2022-2061 with >1,000 40-year soil-moisture sequences from our reconstruction and holding future human-caused drying at today’s level, the drought survives a 23rd year in 94% of sequences.
23/ Regardless whether this drought ends in 1 year or 10, tree-ring records tell us that naturally occurring wet-dry variations will continue. These variations will be super-imposed on an increasingly dry baseline state, however, making future megadroughts increasingly likely.
24/24 This study is an update of a previous analysis of 2000-2018 that was published in @ScienceMagazine. science.sciencemag.org/content/368/64…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
