An acquisition strategy for the proposed NISAR mission to retrieve 3-D

forest structure

Marco Lavalle, Jet Propulsion Laboratory, marco.lavalle@jpl.nasa.gov (Presenter)
Paul A Rosen, Jet Propulsion Laboratory, parosen@jpl.nasa.gov
Scott Hensley, JPL, scott.hensley@jpl.nasa.gov

The proposed NASA-ISRO Synthetic Aperture Radar (NISAR) mission plans to acquire 12-day repeat-pass polarimetric-interferometric radar (PolInSAR) data at nominally zero baseline. If collected with longer spatial baselines, these data could be used to infer forest structure through volumetric decorrelation, and forest dynamics through temporal decorrelation. Forest structure estimation is achieved by comparing polarimetric-interferometric coherence observations against predictions from a physical model, namely the random-motion-over-ground (RMoG) model. Optimal coherence observations are realized with increasing spatial baseline length and at multiple temporal intervals in order to increase sensitivity and robustness of retrieval over a broad range of forest structures. With NISAR, due to the expansive acquisition strategy, the potential to exploit these techniques to produce wall-to-wall global maps of tree height and other structure metrics exists. What needs to be understood is whether the accuracies and resolutions of tree height and structure estimates are amenable to improved estimation of biomass and disturbance compared to other methods, as these are the required science products for NISAR.

We have developed and demonstrated the RMoG-model-based algorithm of tree height estimation through several airborne experiments, including L-band UAVSAR data acquired in Harvard Forest (MA), P-band ONERA’s TROPISAR data acquired over a tropical forest in French Guiana and DLR’s ESAR data acquired over a boreal forest in Sweden. At the workshop, we will describe the algorithm for tree height estimation and show examples of successful retrieval of structural and dynamic forest parameters. We will then move from airborne demonstration to spaceborne implementation, and discuss a range of acquisition scenarios for NISAR, from the exact-repeat, zero-baseline, case to larger baselines, and discuss the performance trades each implies: Since two NISAR science disciplines require short baselines for the mission, any long baseline acquisitions would by necessity be of limited duration. We will evaluate the relative merits of many short baseline acquisitions versus a few long baseline acquisitions, and compare the results to NISAR's current requirements for biomass and disturbance.

Presentation Type:  Poster

Session:  Theme 3: Future research direction and priorities: perspectives relevant to the next decadal survey   (Mon 4:30 PM)

Associated Project(s): 

• Lavalle, Marco: Extracting Structural and Temporal Parameters of Forests from Repeat-Pass Pol-InSAR Data ...details

Poster Location ID: 204