I used to work at JPL, but left in 1998 to work at NASA-Dryden Flight Research Center first as an calibration/validation engineer. Now I work as an Airborne Science Facility Sensor Manager. I am the liaison (a coordinator) between the science community, engineers, and aircraft people bringing in new sensors onto NASA aircraft. I help the scientists perform calibration/validation of their sensors by supplying them with field data collected here at Dryden on Rogers Dry Lake. I have also been trained as an ER-2 mission manager in charge of day-to-day operations of an aircraft during a science campaign [Fig. 7].

Figure 1. We launch radiosondes from a high altitude playa (dry lake), Lunar Lake, NV, to measure atmospheric water vapor over our calibration targets.

I still do some applications work in remote sensing involving oceanography and limnology (study of lakes). I created an algorithm that shows the spatial distribution of mercury contamination in a freshwater lake [Fig 8]. Current projects that I am working include detection of plastic land mines and buried hazardous material, and developing an algorithm that may show hydrogen sulphide emissions on the coast of Namibia, Africa using remote sensing data. My job is really cool as I get to travel (Brasil, Africa, and the US) doing calibration/validation work and I help other scientists by ensuring their data is "good data."

With calibrated and validated data we can be sure that surface changes that we measure are real and not due to the instrument. Calibration/validation is a two-step process of determining the accuracy of information produced from an instrument's data. Calibration concerns the instruments accuracy. No measurement is perfect; consider the measurement of a piece of wood. We can state a length, but we can only know that length to an uncertainty (or accuracy) of about a millimeter, sometimes more if the wood was cut carelessly.

Figure 2. At Rogers Dry Lake, Edwards Air Force Base, CA, which is another calibration target. I am standing behind an "old" manual sun photometer which measures the thickness of the atmosphere by recording how much light reaches the surface.

More specifically, calibration is the process that quantifies the accuracy of the conversion of raw data from the instrument into a property of the object being observed. The second step of validation is the process of making in situ, or ground-based, measurements. These are then compared to measurements of the same area by airborne or spaceborne sensors. This means that when we get certain values (provided all instruments used are calibrated), we know what sort of surface it is 'looking at.' Validation measurements explain what the sensor 'sees.' Every time we launch a new satellite we have a calibration/validation period to make sure we are providing accurate information about Earth. We use aircraft and earth-orbiting satellites to measure geological, biological and oceanographic properties because when platforms are located high above Earth, they give us a very efficient way to get data over large areas. Most satellite data is digital, and these data are different from photographs because the instruments record measurements, such as the amount of light, as a number for each location on the surface. These numbers, which we can view on computers as colors, relate to a physical property such as surface reflectance. Because the data are taken remotely, that is at some distance from the Earth's surface, the measurements of the surface are affected by water vapor and aerosol particles such as dust and smoke. To get accurate information about Earth's surface, we need to remove the effect of the atmosphere and its contents. One of the steps in calibration and validation is to remove these atmospheric effects.

Figure 3. Setting up equipment at Rogers Dry Lake. The radiometer on the left measures surface reflectance. You can see that the playa is a very bright target.

To calibrate/validate sensors, I travel to locations where the remotely sensed data is being acquired. At these locations I take measurements of atmospheric characteristics such as water vapor [Fig. 1] and the thickness of the atmosphere [Fig. 2], as well as measurements of surface reflectance [Figs. 3 & 4]. I make these measurements at the same instant in time as the remote instrument is collecting data. I have collected data at many sites in the Western U.S. including Tomales Bay, Monterey Bay, Lake Tahoe, Mono Lake, Lake Powell. I have also collected data in coral reef areas in Florida. Many calibration studies are done in the desert [Fig. 5] because playas (dry lake-beds) are large, fairly homogeneous bright targets that are easy to locate on satellite and airborne data. In addition, the atmosphere at these sites is easily characterized. Once I am back at JPL, I work with my in situ measurements and the remotely sensed data [Fig. 6]. Using a computer, I develop an algorithm that relates the measurements that were taken by the remote instruments to the actual situation on the surface. An algorithm is a series of arithmetic steps or equations that allows us to take the remotely sensed data and produce numbers that represent some characteristic of the surface. The algorithm contains things like the height of the instrument from the ground and the amount of water vapor in the air. It is very important to calibrate a sensor because it allows us to compare data from different geographical areas or seasons.

Figure 4. This is a “modern” automated sun photometer which ‘tracks’ the sun with no help from me. It is powered by a solar panel.

Figure 5. Working in the dunes at White Sands National Park, New Mexico. Many satellite sensors are calibrated here as the gypsum sand dunes are one of the brightest targets on earth. The All Terrain Vehicle is pulling our bicycle tripod carrying the radiometer. Because it is a national park, we had to sweep over our tracks to leave the surface like we found it.

Figure 6. This is a three color composite image of Rogers Dry Lake, a calibration target. The playa is approximately 30.5 km long, 10 km wide. There are many runways here for landing the Air Force fighter and transport planes. The Space Shuttle also lands here when the weather is bad in Florida.

Figure 7. This is Dennis from California Fish and Game drawing samples that will be analyzed for oxygen content from a CTD onboard the Research Vessel Roger Revelle. His measurements will help me with my chlorophyll algorithm development after I calibrate the Airborne Ocean Color Imager (AOCI).

Figure 8. The spatial distribution of mercury contamination in the freshwater lake, Clear Lake, CA.