NEW!
Science Scope

The Article

The Activities

The Links

Home

Site Map

 

Eyes In The Sky: The Big Picture

by Lisa Young

*Note* The article that appears in Science Scope magazine and the new Science Scope section of this website is based on the following article. They are not identical.

The following article might read like a science fiction script. But it explains how scientists from various disciplines use satellites to study everything from tiny ocean organisms to global climate change. In this story, a research crew that uses high-tech sensors mounted on satelllites gathers information about Earth without physically contacting it. Satellite ocean technology has provided scientists with "eyes in the sky" that enable them to see the water world we call Earth in a whole new light. Join us in our mission to explain how oceanographers get "the big picture."

The Mission: To collect information about the oceans, atmosphere, and land using data from an assortment of satellites orbiting Earth. A crew of scientists must address five major areas of oceanographic research: El Nino, algal blooms, storms and sea level rise, natural disorders, and coral reef ecology.

(If you've been following the Making Waves series, these topics should sound familiar. Hit the Making Waves web site for a refresher!
http://waves.marine.usf.edu)

The primary goal of the mission is to enlighten the readers and crew about the versatility of satellite ocean technology.

The Oceanography Crew:
Sandy Bottms, geological oceanographer
Fisher Reese, biological oceanographer
Dyna McCirculation, physical oceanographer
Sally Nitty, chemical oceanographer

The Guidelines: Only remotely sensed data can be used to accomplish this mission.

The Electromagnetic Spectrum

Gamma rays, microwaves, and radar are terms we use to divide up the continous spectrum of electromagnetic radiation. The spectrum runs from high energy, short wavelength (gamma rays and x-rays) to low energy, long wavelength (TV and radio waves); visible light, the onlypart of the spectrum that humans can see, falls roughly in the middle of these extremes. Each different region of the spectrum has unique properties and uses. Here is a list of the various regions and some common remote sensing applications:

Gamma Rays: Completely absorbed by the upper atmosphere. Not available for satellite remote sensing.
X-Rays: Completely absorbed by the upper atmoshere. Not available for satellite remote sensing.
Ultraviolet: Ozone layer studies
Visible: Estimating ocean phytoplankton bio?, monitoring red tide, ocean circulation studies
Infared: Determining atmospheric and sea-surface temperatures.
MIcrowave: Determining sea-surface height, sea ice studies
Radio: Terrain mapping, hurricane tracking

What is Remote Sensing?
Satellites are "remote sensors" that collect information about the planet without physically contacting it. Everyday examples of remote sensors include our eyes and cameras. A satellite sensor works like our eyes do, with a few important exceptions. Our eyes only collect images in the visible portion of the electromagnetic spectrum (EM spectrum) whereas satellite sensors can collect images from all over the EM spectrum.

Individual sensors collect images at specific regions of the EM spectrum called channels; the channel used depends upon the information desired. For example, to collect plant bio? data, the sensor would use the channels within the visible and infared. For sea surface height data, microwaave channels are used.

When EM radiation (light) hits the surface of an object, different wavelengths are either reflected or absorbed depemding on the physical and chemical properties of the object. The EM radiation that reflects off the objects and returns to the sensor is said to have a particular spectral signature. For example, phytoplankton (tiny ocean plant-like organisms) reflect mostly green light and absorb blue and red light (like leaves on trees). So even though we often can't see phytoplankton with the naked eye, we can use satellites to map areas of high productivity (lots of phytoplankton).


Here are the reports, broken down by topic:

1. El Nino

Dyna McCirculation: I used sea surface height data from the TOPEX/Poseidon satellite to monitor the progression of the 1997-1998 El Nino in the Pacific Ocean.

The height of the sea surface corresponds in part to the temperature of the water; warmer water is "higher" while colder, denser water is "lower." In these images, areas of red and white indicate sea level height well above average (water that's warmer than usual); purple and violet indicate below-average sea level height (and colder water).

The April 1997 image shows an area characterized by above-average sea level height getting sloshed along the equator toward the Americas by east-blowing winds.

The June 1997 image shows that the warmer (higher sea level) water traveled across the Pacific and started spreading along the coasts of North and South America.

In November 1997, El Nino peaked; sea level reached as high as 35 cm above average in parts of the eastern Pacific. Above-average sea level conditions were evident as far north as Alaska and as far south as Chile.

By February 1998, El Nino began to dissipate as the warmer (higher sea level) water receded toward the equator.

By May 1998, the eastern Pacific was returning to normal. However, the western Pacific had not yet returned to normal; the colder (lower sea level) water, which lingered for several months, affected weather conditions around the world.

 

Fisher Reese: The warm waters that encroach upon the west coast of the Americas during an El Nino event significantly impact the food chain, all the way up to humans. Normally, upwelling along the Pacific coast brings cold water and nutrients to the surface; the nutrients support a healthy population of phytoplankton, the tiny ocean "plants" that form the base of the marine food chain. Two images illustrate cold water upwelling and phytoplankton productivity along the west coast of the United States: the Advanced Very High Resolution Radiometer (AVHRR) image shows sea surface temperature, and the Coastal Zone Color Scanner (CZCS) image indicates phytoplankton pigment concentration (which corresponds to productivity). In the AVHRR image, purple and blue denote cold water; in the CZCS image, red and yellow indicate high phytoplankton production. Cold water and high productivity typically characterize the Pacific coast. However, this pattern is disrupted during El Nino, and fish harvests drop dramatically. Satellites help us better understand and predict phytoplankton productivity so we can improve how fisheries are managed.

 

2. Harmful Algal Blooms

 

Fisher Reese: First, here's some background information to skim before I explain how I used satellites to study harmful algal blooms. Algal blooms are natural phenomena that occur when a combination of factors such as sunlight, nutrients, and water temperature encourage the growth of one particular species of phytoplankton over others. Algal blooms are considered harmful if they produce toxins that adversely affect sea life or humans. For example, Gymnodinium breve, a common bloom species along the west coast of Florida, produces a neurotoxin that can contaminate shellfish and make beachgoers cough. Anoxia (oxygen depletion) can also lead to massive fish kills associated with harmful algal blooms. Impacts on public health, local ecology, and tourism underscore the need to study and monitor harmful algal blooms - and satellites can help.

Disadvantages of Satellite Sensors:

* Satellite sensors can only "see" the surface of the water. We can only surmise what is going on below.

* Only a small percentage of the original light remains after the long journey from the sun, through the atmosphere, into the ocean, and back up to the sensor. This means our satellite instruments and the equations we use to understand sensor information must be precise so we don't misinterpret the information.

* To correctly interpret data from satellite sensors, we must compare these data to "ground-truth" data. For instance, before phytoplankton pigment concentration can be derived from SeaWiFS data, we first have to measure how EM radiation changes as it interacts with bodies of water with known quantities of phytoplankton pigment. Then we can develop models that tell us how to interpret the signals we get from satellite sensors.

 

Algal blooms are often called "red tides," brown tides," or "yellow tides" because the abundant phytoplankton actually color the water. Therefore, satellites that measure ocean color, such as the Sea-viewing Wide Field Sensor (SeaWiFS) and CZCS (the predecessor of SeaWiFS), are ideal for detecting and monitoring blooms. This 1978 image of a G. breve bloom off the southwest coast of Florida is one of the first ever obtained by CZCS; this image is an "oldie" but a "goodie." The red, orange, and yellow areas indicate high phytoplankton concentration. Using satellite data, I can determine the concentration of the phytoplankton and measure the extent of a bloom, which are tough to measure from a ship. However, I still need to make initial measurements on land to confirm what the satellites are telling me.

View the image here.

Sally Nitty: I can use satellite sensors to detect chemical markers that signal the end of an algal bloom. Phytoplankton populations can decline for several reasons including: the cells run out of nutrients, zooplankton eat them, sunlight diminishes, or water temperature changes. As phytoplankton die, their cell structure breaks down; "dead materials" are released that give us different signals than live phytoplankton do. The old CZCS sensor was not sensitive enough to make a distinction between these different signals. Fortunately, the new SeaWiFS sensor measures light at the wavelengths that do reveal the difference so we don't overestimate productivity and misinterpret the signals.

3. Storms and Sea Level Rise

Dyna McCirculation: Satellite sensors are lifesavers when it comes to hurricanes. Sensors help us to determine the strength and speed of a hurricane and, more importantly, to make predictions about the path of a hurricane. With advance notice, people who live in the path of a storm can protect themselves and their property, and if necessary, evacuate to a safe location.

View the image here.


In August 1992, Hurricane Andrew ripped through the Bahamas, Florida, and Louisiana. It was one of the most destructive hurricanes in United States history, not to mention the most expensive on record. Damages in the United States exceeded $25 billion. Considering the destructive force of Andrew, the direct loss of life was remarkably low: three in the Bahamas, and 23 died in the United States. The combination of preparations and evacuations minimized the death toll.
In 1954, before Doppler Radar and satellite storm tracking systems, a hurricane similar to Andrew hit the United States with far more deadly results. On October 15, 1954, the weather forecast for the Carolinas called for strong winds and rain in the morning that were to clear in the afternoon. Instead, Hazel, a Category 4 storm like Andrew, made landfall at Myrtle Beach, South Carolina and caught residents off guard. By the time Hazel fizzled out in Canada, it had killed around 350 people. So remember, those spectacular images of hurricanes that you see on the evening news are more than pretty pictures. They can be lifesavers.

4. Natural Disasters

Sally Nitty: Oil spills might be considered more of an "unnatural" disaster, but their impact on the environment is real nevertheless. Surprisingly, major spill events make up just a small percentage of oil pollution overall, but they can be devastating. Satellite imagery is often indispensable in the case of major ocean oil spills such as tanker accidents; we use it to measure the extent of a spill and predict its movement.


This Synthetic Aperture Radar (SAR) image shows an oil spill along the coast of Portugal. The image was taken two days after the Panamanian oil tanker, Cercal, hit a rock and spilled about 1,000 tons of crude oil into the Oporto harbor. The oil quickly spread offshore. The city of Oporto is shown as a cluster of white dots on the coast; that's where the spill began. The oil appears much darker because oil is slick and scatters back less light than the rough sea surface. Sensors that use wavelengths other than radar can also detect oil slicks. However, in this case the SAR instrument was ideal because it operates well in the rainy, foggy weather that characterized the day of the oil spill. Its radar waves pass through clouds uninterrupted, unlike visible or infrared waves.

Sandy Bottoms: I used three Landsat images to study the effects of the eruption of Mount St. Helens in southwest Washington on May 18, 1980. View the images here.

 

The first image shows Mount St. Helens about seven years before the eruption. The white area of the image signifies snow and ice surrounding the mouth of the volcano. The rest of the image is red, which indicates forest vegetation.

 

The next image, taken a few years after the eruption, shows the extent of the devastation. The top 1,300 feet of the once mile-high mountain was lost in the landslide of rock and debris. The grayish-blue areas represent more than 150 square miles of devastated forest that are now mud and ash.

 

The final image, obtained about a dozen years after the eruption, shows the forest gradually making a rebound (light red).

5. Coral Reefs

Sandy Bottoms: Coral reefs are typically found in shallow, clear water. They can often be viewed from space with sensors that use visible and near infrared wavelengths.

This Landsat image shows the southwest portion of Puerto Rico. The offshore reefs are clearly visible (shown in pink/green). We can use satellite images to map reefs, monitor changes in sea-bottom habitats, and look for signs of disease or trauma in the reef community.

Fisher Reese: Many of the organisms that live on a reef fluoresce; that is, they absorb light and radiate it back out at a longer wavelength. In the following photographs, various reef creatures fluoresce particular colors (visible light) after exposure to a flash of UV light (shorter wavelength than visible light).

View the images here.

 

The color varies depending upon the organism, allowing scientists to identify species of coral and other reef organisms based upon their fluorescence signature. A fluorescence sensor mounted on an unmanned underwater vehicle identifies species and creates bottom maps far more quickly and comprehensively than divers could survey the area. With a little more ground work, we may soon be able to use satellites to map larger areas, which would enhance our ability to monitor reef health and study species dynamics.

Conclusion:

These reports highlight just a few of the amazing capabilities of satellite sensors. There are significant advantages to using satellites for ocean research; the two most important are repeatability and coverage. A satellite sensor can regularly repeat its measurements day-in and day-out for many years. The sensor's ability to track information over long periods allows us to assemble time series data to answer questions about weather patterns, productivity cycles, pollution, and climate change. Like a child's growth chart, time series data help us see patterns, make predictions, and develop long-term plans. Satellites also enable us to measure large areas over short periods of time. Some satellites can cover an entire ocean in less than an hour, or even the entire globe in just one day! In contrast, we could send out thousands of ships at once, and many days later would still have only sampled a tiny portion of the oceans. Thanks to satellites, our "eyes in the sky," we have a much clearer picture of the workings of the water world we are so fortunate to call home.

Natural Disasters El 
Nino Oceans From Space Breaking 
News Real Time Data Red 
Tide Sea 
Level Rise Coral 
Reefs