The seeds to a “story”

Over the last few months, we have carefully calculated the historical extreme climate indices specific to the Chesapeake Bay region. Our viewpoint has always been to assess the extreme climate changes and variability rather than the annual mean trends.

Organisms see the day to day variability, not an average.

Figure 1: Excess nutrients can have negative consequences in Chesapeake Bay, such as nuisance algal blooms. Read more here.

Figure 1: Excess nutrients can have negative consequences in Chesapeake Bay, such as nuisance algal blooms. Read more here.

For example, the average temperature for the past 3 days was 79°F…that’s a pretty hot temperature, but not unbearable! But, you do not experience the average temperature, rather, the variability throughout the day. Instead, you want to know: what is the high temperature (91°F) and the low temperature (70°F). That high temperature, on September 9th around 1pm, is the temperature you likely dressed for (tank top weather!).

We hope to demonstrate the utility of these extreme climate indices on specific organism thresholds, such as eelgrass, and other environmental events, such as the transport of nitrogen into the estuary.

Disclaimer: Below is some very initial beginnings to an analysis. My goal was to investigate: is there any relationship between our extreme precipitation indices and the load of total nitrogen in the Patuxent River?

Step One: Use a Corrplot to quickly get correlations

The Corrplot package in R is a great way to visualize which indices are worth pursing!

Here is the data to “tested” for annual-based trends:

Extreme Climate Indices: All annual-based extreme climate indices calculated for the HadEX2 dataset

Figure 2:

Figure 2: The correlation matrix for annual based extreme climate indices.

Streamflow: Annual average streamflow (cubic feet per second) for the Susquehanna, Potomac, James, and Patuxent Rivers. Note that Jug Bay is situated on the Patuxent.

Nutrients: Total nitrogen and total phosphorus (pounds per year) from the Patuxent River.

So which correlations are intriguing (Figure 2)? I will stress that correlation does not mean causation, so just because we see a correlation does not necessarily mean that are connected by cause and effect!

Since my intention was to focus on precipitation, let’s take a closer look at the relationship between total nitrogen and R10mm (the annual count of days when at least 10mm precipitation fell).

A closer look at nitrogen and wet days in the Chesapeake region

Figure 3:

Figure 3: Linear regression of the annual mean nitrogen load with the number of wet days.

The annual mean load of total nitrogen was significantly correlated to the annual number of wet days, where a wet day is defined by at least 10mm of precipitation.

In other words, more total nitrogen was transported by the Patuxent River in years with a high frequency of wet days. This makes perfect sense since runoff, mostly from precipitation, is one of the primary way nutrients are transported from land into Chesapeake Bay.

Figure 4:

Figure 4: Linear regression of annual mean streamflow in the Patuxent River and the R10mm index.

The next “piece” of information I was interested in: is there any relationship between the number of wet days and the annual mean streamflow in the Patuxent River (Figure 4)?

Yes, this seems like an obvious relationship. However, the traditional view has been to look at the amount of rainfall, not necessarily the frequency of wet days. In this way, we provide a new insight to how changes in the amount of rainy days may affect the transport of nutrients.

As expected, but now seen graphically, the more wet days we have each year, the more streamflow.

Figure 5:

Figure 5: The linear regression between total annual precipitation and the number of rainy days.

And because I was curious, how does the total annual precipitation match up against the amount of wet days (Figure 5)?

Turns out they are significantly correlated! Loosely (speculation alert!), this suggests that some years with a high amount of rainfall is due to a high frequency of rainy days. The R10mm index would incorporate high precipitation events, but this relationship shows us that including smaller rain events do matter!

What are the historical patterns of R10mm?

Okay, so my plots above show that years with more rainy days is somewhat related to years with a high nitrogen load in the Patuxent River. So, has this R10mm index historically changed?

Figure 6:

Figure 6: The 21-year rolling mean for the R10mm index in this study area.

Figure 6 is the 21-year rolling mean for 20 different data sets in the Chesapeake region. A few things to notice: 1) All data sets have a general positive slope, 2) but this pattern has a “trough” centered between the 1960 and 1970’s, likely due to a multidecadal teleconnection, and 3) while there is spread between each station, there does not appear to be a latitudinal dependence.

The slope for the HadEX2 dataset suggests that the historical rate of changes for rainy days has been 0.21 days per decade. Note that the HadEX2 data set is a gridded product, meaning that it is an area-weighted, thus a dampened trend.

While a ~0.2 days/decade increase in rainy days may not appear like a lot….it is still 2 day increase as the new normal over the past century.

And what about that variance, or spread in the data?

Figure 7:

Figure 7: The PDF between the 1951-1980 and 1981-2010 climate normals for the R10mm index.

The probability distribution function for the Northern NCDC-Daily stations tells us a lot (Figure 7)! First, there is a rightward shift in the spread of R10mm values. This shift is weighted more towards the right; for example, the 90th percentile from 1951-1980 was 44 days while it is now 47 days in the current climate normal from 1981-2010. That is a +3 day increase between climate normals!

The other detail worth noting is the general broadening of the 1981-2010 probability distribution function. This indicates that, while there is a shift towards more rainy days, there is a chance to have a year with few rainy days as well as a high amount of wet days.

This is where the variability in our study takes a front seat. Yes, we have observed more wet days each year…but that does not mean we will only see years with a high amount of rainy days!

The investigation continues!

Kari Pohl

About Kari Pohl

I am a post-doctoral researcher at NOAA and the University of Maryland (Center for Environmental Science at Horn Point Laboratory). My work investigates how climate variability and extremes affect the diverse ecosystems in Chesapeake Bay. I received a Ph.D. in oceanography from the University of Rhode Island (2014) and received a B.S. in Environmental Science and a B.A. in Chemistry from Roger Williams University (2009). When I am not busy being a scientist, my hobbies include running, watching (and often yelling at) the Boston Bruins, and taking photos of my cat.
This entry was posted in Data, Precipitation, Streamflow and tagged , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *