18 November 2019, Blog No. 7
Every Thursday an updated map of the United States Drought Monitor (USDM)1 is posted at droughtmonitor.unl.edu. A collaborative effort between the USDA2, NOAA3, the National Drought Mitigation Center4, and the University of Nebraska-Lincoln5, the USDM reflects the presence and severity of drought condtions across the United States, Puerto Rico and the U.S. Virgin Islands, and the U.S. Affiliated Pacific Islands (American Samoa, Micronesia, Palau, the Marshall Islands, and the Mariana Islands).
All of these entities use a plethora of resources to best depict hydrological conditions across these regions, and they perform this informed and extensive analysis of drought conditions every week. One of the primary tools used to determine the presence and/or severity of drought conditions is the Palmer Drought Severity Index6 (PDSI or PDI). Developed by meteorologist Wayne C. Palmer7 in 1965, the PDSI is an analytical index that combines rainfall and temperature data among other factors, and serves as a tool to both quantify and qualify long-term drought. The PDSI is rather complex, so for the scope of this blog, just remember that positive PDSI means wetter than normal conditions, and negative PDSI indicates conditions that are drier than baseline.
Figure 1. Time series showing monthly PDSI values for North Carolina from January 1895 to December 2018. Green shading indicates positive PDSI values, indicating wetter than usual conditions, and the brown shading denotes negative PDSI, or drier than normal conditions. Data from NCEI8.
It goes without saying that drought is a multi-faceted natural disaster than has far-reaching impacts across agriculture, society, and the economy. One can glean from the above data that extreme drought is fairly rare, but when it does hit, it often hits hard - persisting for months on end and can even be prolonged to span several years if meteorological conditions yield excessively dry conditions for extended periods.
This blog focused more on the statistical presence of drought conditions in the 20th century and 21st century thus far in North Carolina and does not even scratch the surface of the impacts of drought and the massive scope of drought monitoring. A great appreciation is in line for the hundreds of scientists, hydrologists, and master gardeners that tirelessly pore over data to keep us informed of when drought conditions may be imminent. So, thank you to all of those mentioned and those unmentioned that all keep the well-oiled machine of United States drought monitoring running!
Below is a list of drought monitoring resources and other informative websites that will get you up-to-date on all things drought:
North Carolina Drought Management Advisory Council: https://www.ncdrought.org/
U.S. Drought Portal: https://www.drought.gov/drought/
Drought in North Carolina: https://www.drought.gov/drought/states/north-carolina
Current Drought Conditions and Maps: https://www.drought.gov/drought/data-maps-tools/current-conditions
Drought Impacts by State: https://droughtmonitor.unl.edu/Data/StateImpacts.aspx
USDM Drought Classification Scheme: https://droughtmonitor.unl.edu/About/AbouttheData/DroughtClassification.aspx
USDM Time Series Viewer: https://droughtmonitor.unl.edu/Data/Timeseries.aspx
United States Drought Monitor:
NCEI Climate at a Glance - Statewide Time Series (NC):
Climate Prediction Center:
Figure 2. Legend of PDSI values and their characteristics as listed by the Climate Prediction Center9 on their drought monitoring page, https://www.cpc.ncep.noaa.gov/products/monitoring_and_data/drought.shtml. These PDSI values are assigned to climate divisions in each state across the entire United States.
To the right is a legend of PDSI values screengrabbed from a graphic illustrating PDSI across the nation. The colors are irrelevant for this blog, so just focus on the values - any PDSI value less than -2.0 or greater than 2.0 represent unusual conditions that are exceptionally dry or moist, respectively. PDSI values do occur less than -4.0 and greater than 4.0, although they are seldom recorded. Using PDSI as a proxy, we can look back through monthly data provided by National Centers for Environmental Information to see exceptionally wet and dry periods in North Carolina through the 20th century:
Periods of extreme dry and excessive wetness are of course evident in this data, but what remains astounding to me is the remarkable balance observed in the long-term dataset. In the 1488 months (124 years) of PDSI data, 713 months (47.9%) observed PDSI values greater than zero and 775 of those months (52.1%) observed negative PDSI values. Also, if you find the mean PDSI value from 1895-2019 by averaging all monthly data, we sit at -0.09. In fact, if we look at a histogram showing the distribution of monthly PDSI values, we see a normal distribution with most values in the "near normal" range between -2.0 and 2.0:
Table 5. Table showing the top five longest streaks of abnormally dry conditions in modern North Carolina history (1895-2018). August 1950 - May 1955 was a stretch of nearly five years with drier than normal conditions and thus can be heralded as the driest period in modern North Carolina history.
This blog was designed for a desktop-viewing experience. Viewing on a mobile interface may yield unexpected results.
Tables 1 & 2. The top ten highest (wettest) and lowest (driest) monthly PDSI values in North Carolina from the extended dataset obtained from NCEI. Data ranges from January 1895 to December 2018 and these tables list the top ten PDSI values observed over the entire dataset.
Obviously, with the top ten highest (wettest) PDSI values, nothing would be reflected on the drought monitor - instead, a greater likelihood of flooding occurrences, excessive rainfall, or a lack of dry spells would be found during these times. We can verify this at a glance by realizing that the second highest PDSI value occurred just last December in the wake of Hurricane Florences' flooding rains and subsequent weeks of heavy rain that capped off 2018 as Raleigh's wettest year on modern record. The fall of 2003 was similar with Hurricane Isabel and periodic rains that kept any dryness from persisting in the state. Insufficient data is available to cross-reference for the early 20th century.
We can look back in the USDM Archives10 to find maps (but only since 2000) that correspond to the extremely low PDSI values observed in the fall and winter of 2007 into 2008. Below is an archived map image from the week of 11 December 2007 - one of the most extensive and widespread periods of the most severe degree of drought (D4, Exceptional Drought) in modern North Carolina history.
Figure 3. U.S. Drought Monitor for 11 December 2007. At the release of this drought monitor, 66.2% of North Carolina was in Exceptional Drought (D4), 84.0% of the state was in Extreme Drought (D3), and 100.0% of the state was enduring at least Severe Drought (D2) conditions.
Figure 4. A GIF animation of the development of extreme and exceptional drought across the southeastern United States throughout 2007. All data from UDSM.
The fall months of 2007 and winter months of 2008 that dominate the top-ten lowest PDSI value table above resulted in tremendous agricultural, societal, and economic impacts from the extreme, widespread drought across the entire southeastern United States. Just a quick Google search reveals the broad scope of impacts from the drought of 2007-08. Those search results yield timelines of then North Carolina Governer Mike Easleys' actions to mitigate drought impacts, synoptic analysis that led to drought formation, and it even has its own Wikipedia page.
Figure 5. Histogram of monthly PDSI values from the aforementioned dataset obtained from NCEI. Note the slight bimodal distribution likely due to the PDSI not often yielding values of 0, and the propensity of slightly greater or lesser values to be most common in regular meteorological and hydrological conditions.
When broken down by month and decade, we seen even more consistency and balance between periods of abnormal dryness and wetness:
Tables 3 & 4. Monthly PDSI values aggregated by month and decade or partial-decade. All data analyzed from the same, aforementioned dataset obtained from NCEI. Note that each month in the monthy aggregated data is average of PDSI value from 124 months from 1895-2018. In the decade-breakdown table, the asterisk denotes that data only exists for 1895-1899 and 2010-2018 resulting in incomplete decades for both of those ranges.
In contrast to the balance shown in the tables and plots above, below is a table showing the top five longest streak of negative PDSI values in the 124 year-long dataset used throughout this blog. These streaks of negative PDSI values indicate periods where dry conditions persisted for several months, or even years in the case of these top five streaks. I did not include the top five longest streaks of wet conditions, because they are more numerous and are not as relevant to this particular blog: