Speaker: Simon Zwieback
Remote sensing observations provide an increasingly detailed picture of the Earth’s surface, and they have become indispensable in many branches of geophysics. However, it is not straightforward to isolate the parameters of interest from the raw observations. Systematic errors potentially remain. Nevertheless, these errors are not always explicitly accounted for in applied studies or validation exercises. I will use two examples to illustrate some of the consequences of this neglect. First, systematic vegetation-induced errors in the SMAP soil moisture product are prevalent. Though subtle, they limit the product’s applicability for detecting extreme events and for studying plant-water relations. The potential of systematic errors to give rise to misleading interpretations is also illustrated by the second example, the retrieval of surface deformations from radar interferometry. I will discuss how changes in soil and vegetation water content can bias the deformation estimates, making it difficult to disentangle the complex coupled processes that occur e.g. in permafrost environments. These biases, however, also demonstrate how attention to systematic errors can help improve existing products and spawn off new ones, such as accurate soil moisture estimates from radar interferometry.