Calibration and Validation
Calibration is the set of operations that establish, under specified conditions, the relationship between the values indicated by a measuring instrument or measuring system and the corresponding values realized by standards. Calibration can involve detailed measurements of the instrument hardware and comparison of instrument data products with known ground targets.
Validation is the process of assessing by independent means the quality of the data products derived from the system outputs. NISAR has specific science requirements in each of its disciplines. Validating these requirements entails collecting enough ground measurements or remotely sensed measurements to demonstrate that NISAR's measurements are accurate. For a mission like NISAR with global measurements, the validation plan must incorporate enough global sampling to be representative of expected variability across the globe.
In the context of remote sensing, Cal/Val has become synonymous with verifying that the suite of processing algorithms used to convert raw data into geophysical or biophysical quantities is self-consistent. This can include vicarious calibration, which refers to techniques that use natural or artificial sites on the surface of the Earth for the post-launch calibration of sensors, which is typically called “image calibration” for SAR systems.
Calibration and validation are divided into pre- and post-launch activities. Pre-launch activities focus on instrument calibration, and post-launch activities focus on the data products.
Before launch, calibration and validation activities mainly involve on-ground instrument calibration, algorithm development and evaluation, and establishing the infrastructure and methodologies for post-launch validation. NISAR's algorithm teams will identify requirements for Cal/Val related to specific NISAR data products in their algorithm theoretical basis documents.
The pre-launch objectives of the NISAR Cal/Val program are to:
- Acquire and process data with which to calibrate, test and improve models and algorithms used for retrieving NISAR science data products
- Develop and test techniques and protocols used to acquire validation data and to validate NISAR science products in the post-launch phase.
Pre-launch activities will include development of the calibration procedures and algorithms for the NISAR radar (L1 products), higher-level image products (L2) incorporating such characteristics as geocoding and/or multi-looking, and L3 products, which will be used to validate the NISAR science requirements.
Pre-launch radar instrument calibration will include modeling, analysis, simulations, and laboratory and test-facility measurements. Algorithm development for all products will include testbed simulations, laboratory and test-facility data, field campaigns, exploitation of existing in-situ and satellite data, and utilization of instrument and geophysical models.
At this time, the science team will also identify calibration and validation sites and resources needed for post-launch calibration. For calibration of radar-specific parameters, the project will either deploy or employ existing corner reflector arrays, for example, the array at Rosamond dry lake on Edwards Air Force Base, currently used for calibration of NASA’s airborne L-band radar instrument, the Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR). For some science requirements, ground instrumentation will be deployed to Cal/Val sites before launch and will be verified with contemporary data sources. Contemporary and historical data sets, especially L-band SAR and time series data, will be compiled for Cal/Val sites; demonstration products will be developed for algorithm testing and verification.
After launch, calibration and validation activities will address directly the measurement requirements for the L1 through L3 data products. Each data product has quantifiable performance specifications to be met over the mission lifetime, with calibration and validation requirements addressed in their respective algorithm theoretical basis document.
Post-launch calibration and validation activities are divided into three main phases:
- A 90-day commissioning phase, after which delivery of validated L1 products to the public archive will begin.
- A five-month geophysical product Cal/Val phase, after which delivery of validated L3 products to the public archive will begin.
- The remainder of the science operations phase, in which Cal/Val is performed annually. During this period, additional algorithm upgrades and reprocessing of data products can be implemented if found necessary (e.g., as a result of drifts or anomalies discovered during analysis of the science products), and validation can be done for science requirements that require at least a year’s worth of data.
The post-launch objectives of the Cal/Val program are to:
- Verify and improve the performance of the science algorithms.
- Calibrate or update the calibration of any necessary algorithm parameters.
- Validate the accuracy of the science data products.
The stability of the instrument calibration is verified by continuing to collect calibration data over sites used during the initial 90-day commissioning phase, using the same radar modes as in nominal science operations. This is different from instrument checkout, when multiple modes are used for various calibrations.
The objective of validating the science data products is to assure that global data collected by NISAR meet the project’s L2 science requirements. L3 products will be generated by the science team at the selected validation sites. Validation of the L3 science products will be carried out by a combination of fieldwork and analysis. Each of the three main science disciplines that NISAR supports – solid Earth, cryosphere and ecosystems – has validation sites, networks and platforms detailed in the NISAR Cal/Val Plan (May 2018) (PDF, 30.46 MB).