Test and Evaluation Report Limited Acceptance of the Design Analysis WaterLog® H-3611i Microwave Radar Water Level Sensor.

Abstract

The National Oceanic and Atmospheric Administration (NOAA) National Ocean Service (NOS) Center for Operational Oceanographic Products and Services (CO-OPS) is responsible for developing and maintaining the National Water Level Observation Network (NWLON). CO-OPS, like most operational, technical programs, analyzes state-of-the-art and emerging technologies to identify potential improvements in data quality and operating efficiency and to maintain core expertise for authorized missions. A critical challenge facing CO-OPS is to ensure that water level measuring technologies are capable of delivering data that meet prescribed accuracies, are reliable and resilient in harsh environments, offer improvements in deployment, operation and maintenance efficiencies, and are expected to have a reasonable life-of-industry support for parts manufacturing and service. The ocean observing community has recognized that microwave radar technology, which was previously developed for various range measurement applications, also offers many potential benefits for long-term water level monitoring. In response, the CO-OPS Ocean Systems Test and Evaluation Program (OSTEP) conducted a series of extensive laboratory and field tests on a set of four types of microwave radar sensors from four different manufacturers to determine their suitability for use at NWLON stations and other locations where CO-OPS requires long- and short-term water level measurements observing systems. Analysis of data collected by the selected four sensors over the last 2.5 years of testing points to the Design Analysis WaterLog® H-3611i radar sensor as the best suited for CO-OPS measurement applications at this time. Analysis included an assessment of the four sensors’ water level measurement performance over a broad range of environmental variability. Sensor selection was based on quantitative criteria and a related scoring method specifically designed with CO-OPS’ unique operations and applications in mind. All four sensors demonstrated similar measurement accuracy capabilities, and their scores were very close. However, specific aspects of each sensor influenced the choice of the WaterLog® sensor for this application. Testing of newer versions of the other three sensors, as well those from other manufacturers including Design Analysis, may continue, and they may still be considered for use in CO-OPS operational water level stations. Results presented in this report, however, focus only on measurements collected from WaterLog® radar sensors. Since NWLON sites span more than 200 different coastal locations that are affected by varying combinations of meteorological and oceanographic conditions, field tests of the new microwave radar water level sensor were designed to assess the impact of various environmental parameters on sensor performance. From June to November 2008, test microwave radar sensors were installed at three different NWLON stations with varying coastal environments: Duck, NC; Port Townsend, WA; and Fort Gratiot, MI. Based on analysis of the first year of data from these sites, test microwave radar sensors were installed in 2010 at two additional field test locations: the Bay Waveland, MS and Money Point, VA NWLON stations. Analyses of field results include comparisons between 6-minute (min) average water level measurements collected by the test microwave radar sensor and accepted operating reference NWLON sensors at each site (Aquatrak acoustic at Duck, Port Townsend, Money Point, and Bay Waveland, and BEI float/shaft angle encoder system at Fort Gratiot). In most cases water level measurements from test and operational sensors are in good agreement; however, in some casesmeasurements show deviation closely correlated to changes in environmental conditions. Most notable is the impact of large surface gravity waves (with amplitudes of 1 meter and larger and periods of 10 seconds and longer) and strong long shore and cross shore currents that are most likely set up by wave radiation stress [1,2]. Results from the Duck, NC site, which is an open ocean environment in the most energetic wave regime of the entire East Coast, demonstrate the impacts of the most extreme wave events (significant wave height of approximately 3.5 meters) where monthly WaterLog® versus Aquatrak root mean squared differences (RMSDs) were as large as 7 centimeters (cm), and differences between individual 6-min water level sensors sometimes exceeded 10 cm. Understanding deviations between water levels measured by operational NWLON acoustic sensors and test microwave radar sensors in the presence of a dynamic, open ocean environment such as Duck remains a work in progress; however, observations from the Port Townsend, Money Point, and Fort Gratiot test sites indicate that microwave radar sensors meet accuracy requirements and produce results that generally agree with NWLON sensors. At all three sites, the monthly RMSDs between the Aquatrak and WaterLog® 6-min water level series are generally less than 1 cm, and differences in monthly means are within plus or minus 5 mm. Also, a set of CO-OPS standard water level analysis products generated from an 18-month WaterLog® data record from Port Townsend using the CO-OPS Excel-based Data Management System further confirms that the test sensor can generate accurate measurement results that compare well to those generated by existing NWLON sensors operating in environmental conditions similar to those at the test stations. All test microwave radar data that yielded excellent comparisons with reference NWLON sensors were collected in semi-enclosed, fetch limited, low surface wave coastal environments. Based on these results, OSTEP recommends limited acceptance of the WaterLog® radar as a water level sensor in similar coastal environments. Efforts to facilitate the transition of WaterLog® microwave radar sensors from test to operational status include development of water level quality control (QC) guidelines and a recommended pre-deployment laboratory test procedure specifically designed for this new measurement technology. Extensive analysis of several laboratory tests and 1.5 years of raw 1-Hz data from the Port Townsend field test site were used to optimally tailor previously implemented CO-OPS water level data QC guidelines to accommodate the performance characteristics of the new sensor type. Test results, including problems encountered and lessons learned, have been used to develop and document a standard, four-step microwave radar sensor pre-deployment laboratory test procedure and required data analysis procedures. These laboratory tests are specifically designed to significantly decrease the likelihood of problems during deployment. Although further testing and analysis are needed before a final microwave radar test and evaluation report is issued, most periods of field test data collected by OSTEP to date indicate that microwave radar sensors meet accuracy requirements. Consequently, this report supports operational use of the WaterLog® microwave radar sensor in semi-enclosed, fetch limited coastal regions with a small wave environment.