TOWER-BASED RADAR FOR MONITORING A BOREAL FOREST: MEASUREMENT PERFORMANCE AND DESIGN TRADE-OFFS

In 2016, the BorealScat radar tower experiment was established in a forest site in southern Sweden. The system, based on a 20-channel vector network analyzer, acquires a dense, multi-annual, multi-frequency, multi-polarization tomographic dataset of a forest scene. In this paper we discuss how the system design relates to two performance metrics: backscatter precision and measurement duration. Speckle variance in the time domain is discussed and an approach for avoiding coherent integration loss is presented. These concepts are relevant for any ground-based imaging radar acquiring time series data of dynamic scenes