Remote Sensing Handbook Volume IV - Characterizing tropical forests with multispectral imagery. Chapter 1

Abstract

Tropical forests abound with regional and local endemic species and house at least half of the species on Earth, while covering less than 7% of its land (Lahssini et al., 2022; Bolívar-Santamaría et al., 2021; Bullock et al., 2020; Wilson, 1988; Gentry, 1988; as cited in Skole and Tucker, 1993). Their clearing, burning, draining, and harvesting can make slopes dangerously unstable, degrade water resources, change local climate, or release to the atmosphere as greenhouse gases (GHGs) the organic carbon (C) that they store in their biomass and soils. These forest disturbances accounted for 19% or more of annual human-caused emissions of CO2 to the atmosphere from the years 2000– 2010, and that level is more than the global transportation sector, which accounted for 14% of these emissions. Forest regrowth from disturbances removes about half of the CO2 emissions coming from the forest disturbances (Suab et al., 2024; Bullock et al., 2020; Houghton, 2013; IPCC, 2014). Another GHG of concern when considering tropical forests is N2O released from forest fres. Tropical forests (including subtropical forests) occur where hard frosts are absent at sea level (Holdridge, 1967), which means low latitudes, and where the dominant plants are trees, including palm trees, tall woody bamboos, and tree ferns. They include former agricultural or other lands that are now undergoing forest succession (Zhe et al., 2024; Adrah et al., 2022; Lahssini et al., 2022; FaberLangendoen et al., 2012). They receive from < 1000 MM yr−1 of precipitation to more than ten times that much as rainfall or fog condensation. Whether dry or humid, tropical forests have far more species diversity than temperate or boreal forests, and their role in Earth’s atmospheric GHG budgets is large. Multispectral satellite imagery, i.e., remotely sensed imagery with discrete bands ranging from visible to shortwave infrared wavelengths, is the timeliest and most accessible remotely sensed data for monitoring these forests (Pletcher et al., 2024; Doughty et al., 2023; Ngo et al., 2023; Rana et al., 2023; Bolívar-Santamaría et al., 2021; Cross et al., 2018; Erinjery et al., 2018; Zaki et al., 2017). Given this relevance, we summarize here how multispectral imagery can help characterize tropical forest attributes of widespread interest, particularly attributes that are relevant to GHG emissions inventories and other forest C accounting: forest type, age, structure, and disturbance type or intensity; the storage, degradation, and accumulation of C in aboveground live tree biomass (AGLB, in Mg dry weight ha−1); the feedback between tropical forest degradation and climate; and cloud screening and gap-flling in imagery. In this chapter, the term biomass without further specifcation is referring to AGLB.

Citation

Helmer, E.H., Goodwin, N.R., Gond, V., Souza Jr, C.M. and Asner, G.P.,. 2024. Characterizing tropical forests with multispectral imagery. Chapter 1 (pp 1-44) In: Prasad Thenkabail, Ed., Remote Sensing Handbook Volume IV, 2nd Edition, Forests, Biodiversity, Ecology, LULC and Carbon, CRC Press. 544 p. DOI https://doi.org/10.1201/9781003541172.