ABSTRACT

“Mapping the distribution and abundance of flowering plants using hyperspectral sensing.” This chapter investigates the use of hyperspectral sensing (HS) imaging to map the distribution and abundance of flowering plants as important data sets for landscape ecological and pollination efficacy studies and to prioritize, for instance, weed management interventions. The chapter gives an overview of the spectral characteristics associated with leaf pigmentation changes due to flowering and plant- or canopy-level floral response. The usefulness of spatial data on flowering invasive weeds with regard to their better management and control is highlighted. Important floral detection challenges are discussed, such as mixing effects, intracanopy flowering variability, spatiotemporal variability and patchiness of flowering within or between vegetation communities, and flowering color. Solutions for landscape-based mapping of flowering patterns are discussed. An assessment of important HS-based flower mapping studies is made regarding the thematic depth of the mapping products and the spectral domain and pixel resolution considered in these studies. In general, flowering response is best detected in the visible wavelengths (450–680 nm), while differences in spectral response are due to flowering colors and spectral mixing effects. The assessment showed that for all flowering responses studied with pixel resolutions <2 m, the use of airborne data leads to mapping results with greater thematic depth than results obtained with in situ spectroscopy data. In some studies, airborne very-high-resolution HS data could even be used to map various flowering vegetation communities as well as their flowering colors. Currently there are no spaceborne HS imaging studies on wide-area floral response mapping covering several vegetation types.