Evaluating the Relationship between InterAnnual Climate Variability and Nitrogen
Wet Deposition in the United States
Natalie Thomas1, Tsengel Nergui2, Serena H. Chung2
1University
of North Carolina, Chapel Hill, NC, 2Laboratory for Atmospheric Research, Washington State University, Pullman, WA
Introduction
Correlation Coefficients
a) Correlation between NINO3.4 and Precipitation
c) Correlation between NAO and Precipitation
d) Correlation between NINO and Precipitation -No Lag
Wet deposition rate of nitrogen is governed by precipitation
amount and frequency and thus is influenced by large-scale
climate. This study uses wavelet analysis to determine the
relationship between nitrogen wet deposition and the following
four indices of climate variability.
1 The El Niño Southern Oscillation (ENSO) is a climate
cycle that can be measured by NINO3.4, an index
quantified by the sea surface temperature in the area of 5S5N latitude and 170-120W longitude. El Niño (positive
phase of ENSO) winters are associated with warm, dry
weather in the Pacific Northwest and cool, wet weather in
the Southwest and Southeast US; the opposite is true of La
Niña winters (negative phase of ENSO).
The Arctic Oscillation (AO) is an index defined by
opposing atmospheric pressure patterns in northern
middle and high latitudes. The positive phase of AO
corresponds to low pressure over the polar region and
high pressure at the mid-latitudes, which keeps the eastern
US warmer than normal and brings drier conditions to the
western US.
Data and Methods
• Nitrate (NO3) and ammonium (NH4) wet deposition data and
precipitation data were collected for 79 stations across the
United States from the National Atmospheric Deposition
Program (NADP)1, National Trends Network in the range
of 1979- 2011. Selected stations all had at least 21 years of
continuous data.
• Standardized anomalies were calculated by subtracting
seasonal means from each seasonal value and dividing by the
seasonal standard deviation in order to remove the annual
cycles.
• Climate index data was obtained from the National Oceanic
and Atmospheric Administration’s Climate Prediction
Center2 .
• The wavelet analysis tool3,4,5 was used in order to decompose
one or two time series into both time and frequency space.
0.1
0.34
0.4
0
0
0.45
0.02
0.08
35
0.03
0.49 0.06
0.4
0.29
0.12
0.03
0.18
0.18
30
0.01
0
0.05
0.02
0
0.4
25
0.08
0.04 0.01
0
0
0.01
0.02
0.05
0.16
0
0.05
0
0.02
0.01
0
0.19
0.02
0
0.04
0
0.19
0.22
0.03
0.04
0.01
0
0
0.03
0
0.06
0.02
0.01
0.04
0.04
0.04
0.11 0.04
0.01
0.02
0.08
0.04
0.06 0.18
0.1 0.21
0.22
0.05
0.07
0
0.01
0
0.01
0.08
0.04
0.07
0.03
0.05
0 ≤ R2 < 350.125
0.125 ≤ R2 < 0.25
0.25 ≤ R230< 0.375
0.375 ≤ R2 < 0.50
R2 ≥ 0.50
0
0.03
0.08
0.08
0
0.07
0.01
0
0.17
0.12
0.38 0.03
0.280
0.03
0.36
0.08
0.12
0
0
40
0.04
0.46
0.11
45
0.36
0.32
0.16
0
0.06
0.04
0.2
0.06
0.12
0.19
0.32
0.03
0.04
0.14
0
0.12
0.25
0.02
0.34
0
0
0.04
0.01 0.02
0.140.05
0.05
0.01
0.1
0.03
0.01
0.15
25
-130
-120
-110
-100
-90
d) Correlation betweenLongitude
NINO and Total N Dep -No Lag
-80
-70
-130
b) Correlation between NINO3.4 and Total N Wet Deposition
-120
-110
-100
-90
d) Correlation betweenLongitude
NAO and Total N Dep -No Lag
-80
-70
d) Correlation between NAO and Total N Wet Deposition
50
50
0
0.56
0
0.27
0.22
0
0.02
45
0
0.01
0.24
0.52
0.01
0
0.07
0.31
0.24
0
0.58
0.15
0.18
0.26
0.02
0.07
0.03
0
0.04
Latitude
0
0.13
0.07
0.06
0.29
0.26
0.06
0.06
0.14
0
30
0.05
0
0.01 0.01
0.1
0.07
0.08
0
0
0.06
0.1
0.18
0
0.1
0.03
0.02
0.01
0.03
0.03
0
0.02
0.03
0.06
0
0.1
0.46
0.01
30
0.01
0
0.04
0.12
0
0.04 0.02
0.010.01
0.03
0.01
0.03
0.01
0.03
0 0.01
0.06
0.01
0.2
0
0.12
0.08
0.02
35
0.04
0
0
0.01
0.04
0.14
0.37
0.16 0.04
0 0.07
0.11
0.26
0
0.14
0
0
0.08
40
0.14
0.01
0.01
0.01
0.01
0.02
0.31 0.19
0.350.04
0.13
0.06
0.08
0.12
0.07
0
45
0
0.04
0.01
0.27
0.36
0
0.18
35
0.12
0.27
0.07
0.36
0.02
0.31
0.15
0.01
0
0.08
0.11
0.15
0
0.11
0.01
0.14
0 0.03
0
0
0
0.04
0.2
0.08
0.07
0.29
0.15 0
0.05
0.03
0.01
0.04
0.14
0.5
0.18
25
-120
-110
-100
Longitude
-90
-80
-70
-130
-120
-110
-100
Longitude
-90
-80
-70
Continuous Wavelet and Cross Wavelet Transforms
a) Total N Std.Anomalies, FL41
4
2
0
Normalized anomaly
4
0.5
1
a) Total N Std.Anomalies, FL41
NINO34
Total N Std.Anomalies, FL41
2
NINO34
Total N Std.Anomalies, FL41
0
-2
-4
-2
-4
1980
1980
1985
1990
b) NINO34
1985
1990
0.5
1
1995
2000
1
2
2
4
4
2
4
161980
1980
1985
1990
1985
1995
1990
2005
b) Wavelet Power
1995 Spectrum of NINO34
2000
b) Wavelet Power Spectrum of NINO34
88
2000
2005
1995
2010
Period (years)
0.5
1
2
4
8
16
19800.5
1
1985
1990
1995
8
8
4
2
1
1/2
1/4
1/8
2000
2000
c) Global Wavelet of NINO34
4
1
2
2005
4
2
2010
0
0.2
0.4
2
2
44
1/2
8
1/4
8
1980
1980
1985
1990
1985
1995 1990 2000
2005
1995
2010
20054
0.8
1
1.2
1
2010
0
0.2
0.4
0.6
0.8
1
1/2
1.2
e) Global Wavelet at FL41
1/4
8
1985
1/8
2000
0.6
e) Global Wavelet at FL41
2
d) Wavelet Power Spectrum of Total N Std.Anomalies, FL41
1
1
16
2010
Global Wavelet of NINO34
2005
2010 NINO3.4 andc)Total
Cross
Wavelet Analysis between
Nitrogen Wet Deposition at FL41
c) Total N Std.Anomalies,
d) Wavelet PowerFL41
Spectrum of Total N Std.Anomalies, FL41
Period (years)
(years)
Period
16
Objectives
• To determine the correlation between dominant signals of
climate indices and nitrogen wet deposition through wavelet
transforms.
0.08
0.06 0
0.46
0.07
0.18
0.2
8
• To evaluate the correlation between precipitation and nitrogen
wet deposition
0.27
0.23
40
0.02
0.01
0.17
Latitude
0.12
-130
8
• To identify inter-annual variability of total nitrogen (sum of
ammonium and nitrate) wet deposition
0.23
0.02
Period (years)
4 The North Atlantic Oscillation (NAO) is an index
characterized by differences in pressure between a station
on the Azores and one on Iceland. Positive phases of NAO
are associated with above-average temperatures across the
eastern US.
45
0
0.04
0.04
0.06
0.08
0.02
0.01 0.55
0 0.18
25
Period
(years)
Period(years)
The Pacific/North American (PNA) teleconnection
pattern is associated with above average temperatures in
the western US and below average temperatures in the
southeastern US in its positive phase; the opposite is true
of the negative phase.
0.11
0.03
0.03
0.05
0
0.03
40
Normalized anomaly
3
0
0.13
Figure 1. a) Correlation coefficient (R2) values for NINO3.4
and precipitation; b) R2 values for NINO3.4 and N wet
deposition; c) R2 values for NAO and precipitation; and d)
R2 values for NAO and N wet deposition. All results shown
are without considering lag time.
0.03
0.2
0.1
Period (years)
2
50
0.04
Latitude
Correlations of precipitation and total nitrogen wet
deposition with each of the climate indices were
determined based on the dominant mode of climate
variability. For the ENSO cycle, the coefficients
were calculated using the 2-6 year scale-averaged
variance (band), which was found to be the dominant
periodicity in the 1950-2012 data set based on the
wavelet analysis. The correlations with other indices
were calculated for the 0.5-2 year band. Results are
shown here for both the NINO3.4 and the NAO with
precipitation and combined NH4 and NO3 wet
deposition. The red arrow indicates station FL41,
which is used as an example for the wavelet
transforms below.
50
Latitude
Human inputs of nitrogen into the atmosphere have increased
dramatically as a result of fossil fuel emissions and fertilizer
use. When this excess nitrogen deposits onto the Earth’s
surface, it has many diverse and harmful effects, ranging from
reduced drinking water quality to altered chemical composition
of ecosystems.
d) Correlation between NAO and Precipitation -No Lag
2005
Time (year)
1990
2010
1995
0
2000
0.5
2005
1
1.5
Power
2010
2
1/8
2.5
• The wavelet analysis tool makes it possible to determine the
dominant periodicities of a time series and where in time these
occur. The darker red colors indicate areas of high power, or
areas where the variance is largest. The black contours display
statistically significant variances at the 95% confidence level.
• As seen in the both the time series and the continuous wavelet
spectra, nitrogen wet deposition at FL41 has an area of
significantly high variance around 1997, which corresponds to
the strong El Niño event of 1997-1998.
• The cross wavelet transform reveals that the two time series
share a period of high common power centered around 1997.
• The arrows represent the phase relationship, with right-pointing
arrows representing in-phase and left-pointing representing out
of phase. In this case, the arrows, which point roughly
southeast, indicate that NINO3.4 leads N deposition by about 4-5
months.
Figure 2. a) Time series of NINO3.4 and FL41 total nitrogen wet deposition standardized anomalies;
b) continuous wavelet transform of NINO3.4; c) continuous wavelet transform of FL41 nitrogen wet deposition;
1980and d) cross
1985
1990
2000
2005 wet deposition.
2010
0
0.5
1
1.5
2
2.5
wavelet transform
between1995
NINO3.4 and
FL41 nitrogen
Time (year)
Power
Conclusions and Future Work
References & Acknowledgements
• Results of the wavelet analysis were very site specific, but in general, they reveal that nitrogen wet
deposition has a dominant periodicity of 0.5-2 years; for ENSO it is 2-6 years, and for AO, NAO and PNA,
it is 0.5-2 years in the range of 1979-2011.
• Precipitation and nitrogen wet deposition were most correlated in the Great Plains region, in the states of
South Dakota, Kansas, Iowa, Tennessee and Ohio, with R2 values in the range of 0.50 to 0.75. Lower
correlations (R2 values less than 0.125) were seen in the states such as Montana, Wyoming and Utah.
• ENSO was found to be most correlated with precipitation and nitrogen deposition in the Midwest and
eastern US, with many R2 values in the range of 0.38 to 0.50; smaller values were found for the western
US, but there were very few NADP sites in this region.
• Calculated correlations between NAO, AO and PNA and precipitation and nitrogen deposition were not as
strong as for ENSO, and the strongest of these were again seen in the Midwest and eastern US, with a
maximum R2 value of 0.29.
1. National Atmospheric Deposition Program:
http://nadp.sws.uiuc.edu
2. National Oceanic and Atmospheric Administration, Climate
Prediction Center: http://www.cpc.ncep.noaa.gov/data/indices
3. Grinsted, A., J. C. Moore, and S. Jevrejeva (2004), Application
of the cross wavelet transform and wavelet coherence to
geophysical time series, Nonlin. Processes Geophys., 11(5/6),
561–566, doi:10.5194/npg-11-561-2004.
4. Liu, Y., X. San Liang, and R. H. Weisberg (2007), Rectification
of the bias in the wavelet power spectrum, J Atmos. Oceanic
Technol., 24(12), 2093–2102, doi:10.1175/2007JTECHO511.1.
5. Torrence, C., and G. P. Compo (1998), A practical guide to
wavelet analysis, Bull. Amer. Meteor. Soc., 79(1), 61–78,
doi:10.1175/1520-0477(1998)079<0061:APGTWA>2.0.CO;2.
Future research will include:
• The use of precipitation frequency instead of precipitation rate data in the analysis with N wet deposition.
• Analysis with dry deposition data from CASTNet.
This work was supported by the NSF's Research Experiences for
Undergraduates program (grant AGS-1157095) and by the
USDA grant 20116700330346 NIFA.