Last week, we presented water temperature data to give insights on eelgrass die-off events in Chesapeake Bay and demonstrated that air temperature was tightly correlated to water temperature.
In other words, changes to atmospheric temperature will directly affect the shallow water environment.
In today’s post, we will show data from two extreme climate indices to further investigate the occurrence of thermal threshold exceedances (when a temperature rises above a critter’s biological tolerance). Our goal is to demonstrate the utility of the extreme climate indices in investigating ecological questions, such as the occurrence of eelgrass die-off events when water temperatures exceed 30°C.
The Climate Indices
TX90p: This index is the percentage of days when the daily maximum temperature (TMAX) is greater than the 90th percentile, calculated on a monthly interval. In other words, it is the %days when the daily high is much warmer than average.
To compare what exceeds the 90th percentile, we need to know what the 90th percentile is! That percentile is calculated from the 1961-1990 climate normal.
The TMAX 90th percentile for summer (June, July, and August) is 33.3°C (91.9°F) in Maryland and 33.9°C (93.0°F) in Virginia.
TX10p: This index is the percentage of days when the daily maximum temperature (TMAX) is less than the 10th percentile, calculated on a monthly interval. In other words, it is the amount of days when the daily high is much cooler than average.
The TMAX 10th percentile for summer (June, July, and August) is 25.0°C (77.0°F) in Maryland and 25.6°C (78.1°F) in Virginia.
This data allows us to know the extreme “range” of the daily maximum temperature in the summer months.
The Historical Timeline
How has the TX90p and TX10p indices changed during our >110 year time series? Have they changed? More specifically, is there any trend during the summer months (when eelgrass die-offs are most likely to occur)?
These are the general questions we can speculate with our calculated extreme climate indices.
In Figure 2, we have the TX90p timeseries for all NCDC-Daily stations combined to represent the weather experienced by the near-shore Chesapeake Bay.
One thing to notice: the two years with the highest amount of days in the TMAX 90th percentile (roughly the temperature at which eelgrass may start to suffer) occurred in the last 15 years (2002 and 2010). Additionally, since 1980, the “wiggles” appear to surpass a 10-year event more frequently.
I applied a 21-year rolling mean and variance (Figure 3) to look at the TX90p index for the summer months even closer. The mean suggests that the TX90p index has significantly increased over the last ~110 years in both the North and Southern sections of Chesapeake Bay….with some sinusoidal oscillations that appear in most of our indices (likely from teleconnections and larger climate processes). Interestingly, we also see a sharp increase in the variance starting ~1970.
The percentage of days when TMAX surpasses ~ 30°C appears to have increased. BUT, the variance has also increased. Loosely speaking, the occurrence of an eelgrass die-off is more likely, but these events may be followed by summers with no or little die-off events.
Ecology and weather are complicated!
The TX90p index lets us know how many really warm days we get each summer, but for an eelgrass die-off, we are also curious about those basement temperatures. In other words, are the coolest TMAX temperatures changing?
In Figure 4, the percentage of days below the 10th percentile appear to decrease, meaning we are seeing less cooler days in the summer (in reference to TMAX). Thus, eelgrass (and people!) in Chesapeake Bay are experiencing more days above the 90th percentile and less days below the 10th percentile. Overall verdict, it appears to be getting more common to get warmer days more frequently!