Richard P. Stumpf

Interannual Variability of Cyanobacterial Blooms in Lake Erie

After a 20-year absence, severe cyanobacterial blooms have returned to Lake Erie in the last decade, in spite of negligible change in the annual load of total phosphorus (TP). Medium-spectral Resolution Imaging Spectrometer (MERIS) imagery was used to quantify intensity of the cyanobacterial bloom for each year from 2002 to 2011. The blooms peaked in August or later, yet correlate to discharge (Q) and TP loads only for March through June. The influence of the spring TP load appears to have started in the late 1990 s, after Dreissenid mussels colonized the lake, as hindcasts prior to 1998 are inconsistent with the observed blooms. The total spring Q or TP load appears sufficient to predict bloom magnitude, permitting a seasonal forecast prior to the start of the bloom.

Relating spectral shape to cyanobacterial blooms in the Laurentian Great Lakes

A change in the spectral shape at 681 nm is used to distinguish blooms of cyanobacteria from blooms of other phytoplankton via MERIS satellite sensor imagery. During large cyanobacterial blooms, the spectral shape around 681 nm is not a positive quantity as scattering due to cyanobacteria overwhelms the fluorescence signal, thus creating a negative spectral shape. This relationship is consistent in both remotely sensed and in situ data.

Satellite detection of bloom and pigment distributions in estuaries

Using a form of vector analysis of satellite spectral data, it is possible to distinguish variations in water color and pigment concentrations from changes in turbidity. In turbid water (reflectance > 0.01), the orientation of a spectral vector depends predominantly on organic pigment absorption. Turbidity controls the total reflectance. A reflectance model based on the vector expression leads to a relationship nc ∝ Cji for two bands or nc ∝ Cji0.5 for multiple band analyses, where nc is the chlorophyll concentration and Cji is the vector orientation in bands j and i. With a simple uniform atmospheric correction to AVHRR and CZCS satellite data, maps of Cji (i = red and j = near-infrared bands) show the location of blooms in the Chesapeake Bay in the springs of 1981 and 1982. Regressions of in situ chlorophyll concentrations (nc) with Cji support the described relationships. The atmospheric correction has a potential error of 0–30% for Cji, with the error tending to compress the range of Cji. The scene bias may be < 25%. The bias and pixel errors combined produce an error of about ±60%, although bloom composition may produce some additional error. The results indicate that this method can identify blooms in estuaries where the reflectance is between 0.01 and 0.07 and, with some calibration, may provide estimates of chlorophyll for concentrations greater than 5 μg/L.

Remote Estimation of Water Clarity in Optically Complex Estuarine Waters

Abstract AVHRR satellite imagery was evaluated as a potential data source for monitoring light attenuation (KPAR), as a measure of turbidity, in Pamlico Sound estuary, North Carolina. In situ water quality data and reflectance imagery collected on 10 different dates were used to calibrate a general optical equation relating satellite-derived reflectance (Rd), nominally R(630 nm) to KPAR. Additional spectral data (e.g., absorption, subsurface reflectance), related reflectance and KPAR to changes in phytoplankton pigments, organic matter, and suspended sediments. Optically, Pamlico Sound, North Carolina is dominated by scattering from suspended sediments, whereas the tributary rivers are dominated by absorption from both dissolved and particulate organic matter. A general relationship developed between Rd and KPAR (r2=0.72) in Pamlico Sound was found useful in a variety of environmental conditions; however a relationship between Rd and suspended sediment concentration was less robust, and affected by changing sediment characteristics. In the rivers, high and variable absorption in the visible wavelengths precluded development of a relationship between Rd and KPAR. The relationship developed between Rd and KPAR in Pamlico Sound is similar to those determined for Delaware Bay and Mobile Bay in previous studies, suggesting possible broader regional application of algorithms for coastal bays and estuaries having similar sediment characteristics, with direct application to SeaWiFS data. Published by Elsevier Science Inc.

Wind and tidal forcing of a buoyant plume, Mobile Bay, Alabama

AVHRR satellite imagery and in situ observations were combined to study the motion of a buoyant plume at the mouth of Mobile Bay, Alabama. The plume extended up to 30 km from shore, with a thickness of about 1 m. The inner plume, which was 3–8 m thick, moved between the Bay and inner shelf in response to tidal forcing. The tidal prism could be identified through the movement of plume waters between satellite images. The plume responded rapidly to alongshore wind, with sections of the plume moving at speeds of more than 70 cm s−1, about 11% of the wind speed. The plume moved predominantly in the direction of the wind with a weak Ekman drift. The enhanced speed of the plume relative to normal surface drift is probably due to the strong stratification in the plume, which limits the transfer of momentum into the underlying ambient waters.

Variations in water clarity and bottom albedo in Florida Bay from 1985 to 1997

Following extensive seagrass die-offs of the late 1980s and early 1990s, Florida Bay reportedly had significant declines in water clarity due to turbidity and algal blooms. Scant information exists on the extent of the decline, as this bay was not investigated for water quality concerns before the die-offs and limited areas were sampled after the primary die-off. We use imagery from the Advanced Very High Resolution Radiometer (AVHRR) to examine water clarity in Florida Bay for the period 1985 to 1997. The AVHRR provides data on nominal water reflectance and estimated light attenuation, which are used here to describe turbidity conditions in the bay on a seasonal basis. In situ observations on changes in seagrass abundance within the bay, combined with the satellite data, provide additional insights into losses of seagrass. The imagery shows an extensive region to the west of Florida Bay having increased reflectance and light attenuation in both winter and summer begining in winter of 1988. These increases are consistent with a change from dense seagrass to sparse or negligible cover. Approximately 200 km2 of these offshore seagrasses may have been lost during the primary die-off (1988 through 1991), significantly more than in the bay. The imagery shows the distribution and timing of increased turbidity that followed the die-offs in the northwestern regions of the bay, exemplified in Rankin Lake and Johnson Key Basin, and indicates that about 200 km2 of dense seagrass may have been lost or severely degraded within the bay from the start of the die-off. The decline in water clarity has continued in the northwestern bay since 1991. The area west of the Everglades National Park boundaries has shown decreases in both winter turbidity and summer reflectances, suggestive of partial seagrass recovery. Areas of low reflectance associated with a majorSyringodium filiforme seagrass meadow north of Marathon (Vaca Key, in the Florida Keys) appear to have expanded westward toward Big Pine Key, indicating changes in the bottom cover from before the die-off. The southern and eastern sections of the Bay have not shown significant changes in water clarity or bottom albedo throughout the entire time period.

Spatial and Temporal Patterns in the Seasonal Distribution of Toxic Cyanobacteria in Western Lake Erie from 2002–2014

Lake Erie, the world’s tenth largest freshwater lake by area, has had recurring blooms of toxic cyanobacteria for the past two decades. These blooms pose potential health risks for recreation, and impact the treatment of drinking water. Understanding the timing and distribution of the blooms may aid in planning by local communities and resources managers. Satellite data provides a means of examining spatial patterns of the blooms. Data sets from MERIS (2002–2012) and MODIS (2012–2014) were analyzed to evaluate bloom patterns and frequencies. The blooms were identified using previously published algorithms to detect cyanobacteria (~25,000 cells mL−1), as well as a variation of these algorithms to account for the saturation of the MODIS ocean color bands. Images were binned into 10-day composites to reduce cloud and mixing artifacts. The 13 years of composites were used to determine frequency of presence of both detectable cyanobacteria and high risk (>100,000 cells mL−1) blooms. The bloom season according to the satellite observations falls within June 1 and October 31. Maps show the pattern of development and areas most commonly impacted during all years (with minor and severe blooms). Frequencies during years with just severe blooms (minor bloom years were not included in the analysis) were examined in the same fashion. With the annual forecasts of bloom severity, these frequency maps can provide public water suppliers and health departments with guidance on the timing of potential risk.

Estimating cyanobacterial bloom transport by coupling remotely sensed imagery and a hydrodynamic model.

The ability to forecast the transport of harmful cyanobacterial blooms in the Laurentian Great Lakes is beneficial to natural resource managers concerned with public health. This manuscript describes a method that improves the prediction of cyanobacterial bloom transport with the use of a preoperational hydrodynamic model and high temporal resolution satellite imagery. Two scenarios were examined from separate cyanobacterial blooms in western Lake Erie, USA. The first scenario modeled bloom position and extent over the span of 13 days. A geographic center, or centroid, was calculated and assigned to the bloom from observed satellite imagery. The bloom centroid was projected forward in time, and the projected position was compared to the final observed bloom centroid. Image pixels flagged as cyanobacterial bloom were compared between the initial image and the final image, and this was assumed as persistence. The second bloom scenario was modeled for a period of 12 days, and the results were framed in an ecological context in an effort to gain further understanding of cyanobacterial bloom dynamics. These modeling techniques can be incorporated into an operational forecasting system.

Challenges for mapping cyanotoxin patterns from remote sensing of cyanobacteria

Using satellite imagery to quantify the spatial patterns of cyanobacterial toxins has several challenges. These challenges include the need for surrogate pigments - since cyanotoxins cannot be directly detected by remote sensing, the variability in the relationship between the pigments and cyanotoxins - especially microcystins (MC), and the lack of standardization of the various measurement methods. A dual-model strategy can provide an approach to address these challenges. One model uses either chlorophyll-a (Chl-a) or phycocyanin (PC) collected in situ as a surrogate to estimate the MC concentration. The other uses a remote sensing algorithm to estimate the concentration of the surrogate pigment. Where blooms are mixtures of cyanobacteria and eukaryotic algae, PC should be the preferred surrogate to Chl-a. Where cyanobacteria dominate, Chl-a is a better surrogate than PC for remote sensing. Phycocyanin is less sensitive to detection by optical remote sensing, it is less frequently measured, PC laboratory methods are still not standardized, and PC has greater intracellular variability. Either pigment should not be presumed to have a fixed relationship with MC for any water body. The MC-pigment relationship can be valid over weeks, but have considerable intra- and inter-annual variability due to changes in the amount of MC produced relative to cyanobacterial biomass. To detect pigments by satellite, three classes of algorithms (analytic, semi-analytic, and derivative) have been used. Analytical and semi-analytical algorithms are more sensitive but less robust than derivatives because they depend on accurate atmospheric correction; as a result derivatives are more commonly used. Derivatives can estimate Chl-a concentration, and research suggests they can detect and possibly quantify PC. Derivative algorithms, however, need to be standardized in order to evaluate the reproducibility of parameterizations between lakes. A strategy for producing useful estimates of microcystins from cyanobacterial biomass is described, provided cyanotoxin variability is addressed.

Remote estimation of the diffuse attenuation coefficient in a moderately turbid estuary

Abstract Solutions of the radiative transfer equation are used to derive relationships of water reflectance to the diffuse attenuation coefficient (K) in moderately turbid water (K > 0.5 m−1). Data sets collected from the NOAA AVHRR and in situ observations from five different dates confirm the appropriateness of these relationships, in particular the logistic equation. Values of K calculated from the reflectance data agree to within 60% of the observed values, although the reflectance derived using a more comprehensive aerosol correction is sensitive to chlorophyll concentrations greater than 50 μg L−1. Agreement between in situ and remote observations improves as the time interval between samples is narrowed.