Evaluating historical trends and influences of meteorological and seasonal climate conditions on lake chlorophyll a using remote sensing

Figures & data

Figure 1. Great Salt Lake System, with Utah Lake draining into the Great Salt Lake via the Jordan River. The southern arm of the lake is further divided by a causeway separating Farmington Bay from the rest of the Great Salt Lake. Sampling locations are shown each of the contributing monitoring organizations.

Table 1. Summary data used for model development, application, and analysis.

Data source	Data type	Period of record	Resolution or units
UDWQ	Surface chlorophyll a	1992–2009	µg/L
USGS	Surface chlorophyll a	1995–2015	µg/L
Jordan River/Farmington Bay Water Quality Council	Surface chlorophyll a	2013	µg/L
Landsat-5 TM	Surface reflectance	1984–2007	30 m pixels, 16-d satellite overpass
Landsat-7 ETM+	Surface reflectance	1999–2016	30 m pixels, 16-d satellite overpass
National Climate Data Center	Daily precipitation, daily average wind speed, daily total wind movement, daily maximum air temperature	1984–2016	mm/d, m/s, km/d, degrees Celsius
Natural Resources Conservation Service	Snow water equivalent (SWE)	1984–2016	inches/d

Figure 2. Modeled versus observed Chl-a, with R² between modeled and observed values. Differences between the solid regression line and the dashed 1:1 line show the under-/overestimation.

Table 2. Summary of model characteristics and performance. Performance is measured via R², mean absolute error (MAE), and percent bias (PBIAS) between modeled and observed Chl-a concentrations.

Lake	Season	Number of near coincident matches	Time window (±days)	Range of Chl-a in calibration data (µg/L)	Modeled vs. observed R^2	MAE (µg/L)	PBIAS	Model
Utah Lake	May–Jun	10	1	1.2–45	0.99	0.9	−0.3%	Chl-a = exp(−14.23 + 9.33*Green/Blue + 0.003*Blue − 0.004*SWIR1)
	Jul–Sep	46	1	0.2–185.5	0.88	10.6	−15.9%	Chl-a = exp(7.33 - 0.004*Blue − 0.05*Green/SWIR2 + 0.01*Red/SWIR1)
Great Salt Lake	May–Jun	90	2	0.3–164.2	0.86	10.9	17.0%	Chl-a = exp(3.51 + 0.09*Green/Blue − 0.01*Blue + 0.009*Red)
	Jul–Sep	69	2	0.045–112	0.79	5.2	−3.0%	Chl-a = exp(−0.36 + 6.12*Red/Blue − 0.006*Red)
Farmington Bay	May–Jun	14	4	4–505.9	0.87	38.5	1.4%	Chl-a = exp(6.53 - 1.02*Red/NIR − 0.009*SWIR2 + 0.004*SWIR1)
	Jul–Sep	11	4	37.38–196	0.86	13.4	0.2%	Chl-a = exp(6.11 − 0.002*Green)

Figure 3. Historical record of Chl-a concentrations between 1984 and 2017 for the sampling locations in the Great Salt Lake System: (A) Utah Lake, (B) Great Salt Lake, and (C) Farmington Bay. The observed data (triangles) are from field samples and the estimated (circles) are from remotely sensed imagery.

Table 3. Summary of statistically significant (α = 0.05) long-term trends (change per year), calculated using the Theil–Sen estimator from 1984 to 2016.

Lake	Average Chl-a (µg/L/yr)	Extreme Chl-a (µg/L/yr)	Standard deviation of Chl-a (µg/L/yr)	Bloom extent (%/yr)	Timing of peak average Chl-a (d/yr)	Timing of peak extreme Chl-a (d/yr)
Utah Lake	−0.25	1.8	0.2	−0.2	−1.2	−0.5
Great Salt Lake	—	2.8	0.55	—	−1.5	−2.5
Farmington Bay	—	4.9	1.2	−0.4	−2.5	−1.2

Figure 4. Magnitudes of trends for lake-wide (A) average (mean) Chl-a, (B) extreme (99th percentile) Chl-a, (C) variability (standard deviation) of Chl-a, and (D) bloom extent (as a percent of lake >50 µg/L). Only statistically significant trends are shown. Note increasing trends in early summer months for average and extreme concentrations, increasing trends in variability for the whole growing season, and increasing trends for bloom extent in the early summer months in the GSL.

Figure 5. Correlation between lake-wide bloom measures and meteorological parameters, as measured by Kendall’s tau for (A) Utah Lake, (B) GSL, and (C) Farmington Bay. The only significant correlations are between bloom measures and temperature (positive for Utah Lake, and negative for GSL).

Figure 6. Differences in distributions of Chl-a concentrations for selected sites in Utah Lake, GSL, and Farmington Bay. Chl-a is grouped according to the meteorological conditions of the previous day: (A) precipitation, (B) wind, and (C) maximum air temperature). Note the difference in sensitivity to different meteorological conditions between the lakes and between different locations in the same lake.

Figure 7. Distributions of annual extreme Chl-a concentrations, grouped by the condition of the seasonal climate parameters: (A) winter precipitation, (B) summer precipitation, (C) summer temperature, and (D) SWE. For Utah Lake, differences in extreme values generally follow the conditions suggested by Page et al. (2018); high Chl-a coincides with high winter precipitation, summer temperature, and SWE.

Page BP, Kumar A, Mishra DR. 2018. A novel cross-satellite based assessment of the spatio-temporal development of a cyanobacterial harmful algal bloom. Int J Appl Earth Obs Geoinf. 66:69–81. doi:10.1016/j.jag.2017.11.003.

 Web of Science ®Google Scholar