We address the feasibility of a GNSS-R code-altimetry space mission and more specifically a dominant term of its error budget: the reflected-signal range precision. This is the RMS error on the reflected-signal delay, as estimated by waveform retracking. So far, the approach proposed by [Lowe et al., 2002] has been the state of the art to theoretically evaluate this precision, although known to rely on strong assumptions (e.g., no speckle noise). In this paper, we perform a critical review of this model and propose an improvement based on the Cramer-Rao Bound (CRB) approach. We derive closed-form expressions for both the direct and reflected signals. The performance predicted by CRB analysis is about four times worse for typical space mission scenarios. The impact of this result is discussed in the context of two classes of GNSS-R applications: mesoscale oceanography and tsunami detection.
تحميل البحث