What is it about?

Land use/land cover mapping is usually performed by classifying satellite imagery (e.g., Landsat, Sentinel) for the whole survey region using classification algorithms implemented with training data. Subsequently, probabilistic samples are usually implemented with the main purpose of assessing the accuracy of these maps by comparing the map class and the ground condition determined for the sampled units. The main proposal of this paper is to directly exploit these probabilistic samples to estimate the land use/land cover class at any location of the survey region in a design-based framework by the well-known nearest-neighbour interpolator.

Featured Image

Why is it important?

For the first time, the designbased consistency of nearest-neighbour maps (i.e., categorical variables) is theoretically proven and a pseudo-population bootstrap estimator of their precision is proposed and discussed. These nearest-neighbour maps provide the ability to place mapping within a rigorous design-based inference framework, in contrast to most traditional mapping approaches which often are implemented with no inferential basis or by necessity (due to lack of a probabilistic sample) model-based inference.

Perspectives

The mapping performed by NN interpolation exploits only information arising from space without taking advantage of the knowledge of remote-sensing covariates (i.e., auxiliary variables) often available for the whole survey region. Indeed, incorporating other auxiliary variables is straightforward in design-based mapping of quantitative variables in which the survey variable is predicted by a suitable function of the auxiliary variables and the design-based interpolation is performed on the residuals. However, this approach is less straightforward in LULC mapping because classes are categorical variables. An alternative way to exploit auxiliary information is to perform NN interpolation in the space of auxiliary variables; that is, the interpolated class at a location is the class observed at the sample location that is nearest in the auxiliary space rather than in the coordinate space. This intriguing idea necessitates further theoretical investigations to be fully confirmed, especially concerning the property of consistency.

Piermaria Corona
CREA Research Centre for Forestry and Wood

Read the Original

This page is a summary of: Design-based mapping of land use/land cover classes with bootstrap estimation of precision by nearest-neighbour interpolation, The Annals of Applied Statistics, December 2023, Institute of Mathematical Statistics,
DOI: 10.1214/23-aoas1754.
You can read the full text:

Read

Contributors

The following have contributed to this page