**1. Introduction**

76 Remote Sensing of Biomass – Principles and Applications

Proisy, C., Mougin, E., Fromard, F., Trichon, V. & Karam, M. A. (2002). On the influence of

Proisy, C., Couteron, P. & Fromard, F. (2007). Predicting and mapping mangrove biomass

Rao, A. R. & Lohse, G. L. (1996). Towards a texture naming system: Identifying relevant

Rich, R. L., Frelich, L., Reich, P. B. & Bauer, M. E. (2010). Detecting wind disturbance

Richards, P. W. (August 1996). *The Tropical Rain Forest. An Ecological Study*, 2nd edition,

Vincent, G. & Harja, D. (2008). Exploring Ecological Significance of Tree Crown Plasticity

Zhao, K., S. Popescu and R. Nelson (2009). Lidar remote sensing of forest biomass: A scale-

Environment, Vol( 113), No.1, (January 2009), pp. 182-196, ISSN: 0034-4257

Cambridge University Press, ISBN: 9780521421942, Cambridge

No.2, (February 2010), pp.299-308, ISSN: 0034-4257

ouvertes.fr/hal-00509952/fr/>

pp.1221-1231, ISSN: 1095-8290

ISSN: 0034-4257

ISSN: 0042-6989

India. *Pondy Papers in Ecology*, 10: 1-71, Available from <http://hal.archives-

canopy structure on the polarimetric radar response from mangrove forest. *International Journal of Remote Sensing,* Vol.23, No.20, pp.4197-4210, ISSN: 0143-1161

from canopy grain analysis using Fourier-based textural ordination of IKONOS images. *Remote Sensing of Environment,* Vol.109, No.3, (August 2007), pp.379-392,

dimensions of texture. *Vision Research,* Vol.36, No.11, (June 1996), pp.1649-1669,

severity and canopy heterogeneity in boreal forest by coupling high-spatial resolution satellite imagery and field data. *Remote Sensing of Environment,* Vol.114,

through Three-dimensional Modelling. *Annals of Botany,* Vol.101, No.8, (May 2008),

invariant estimation approach using airborne lasers. Remote Sensing of

Biomass and Leaf Area Index (LAI) are two important biophysical properties of vegetation as they inform about vegetation production. LAI is directly related to the exchange of energy and mass between plant canopies and the atmosphere (Fassnacht *et al*., 1997), while biomass reflects the amount of carbon converted through photosynthesis and accumulated in the different plant components. Thus, the two variables reflect much of the potential and actual production of plant ecosystems (Kasischke *et al*., 2004).

Fire is ubiquitous in most terrestrial ecosystems causing spatial patterning at many scales (Chapin III *et al*., 2003). In tropical savannas in general and, in the southern African savannas in particular, much of the ecosystem functioning is largely defined by the combination of climate, fires and herbivory. Andreae (1993) estimated that fires in the African and the world savannas account, respectively, for 22% and 42% of the biomass burned globally. Moreover, the amount of CO2 exchanged with the atmosphere in southern Africa may represent up to 20% of the regional net primary production (Scholes & Andreae, 2000).

In spite of the elevated importance of disturbances in miombo woodlands, there still is a gap in the understanding of the interaction between them and vegetation. This, results partially from the short temporal and spatial scales of observation of much of the existing studies. For example, except for the long-term experimental study carried out in Zambia for 15 years (Trapnell, 1959), the other studies are all *points* in space and time, much of them lasted less than 5-years. Moreover, they address a specific aspect of miombo woodlands functioning, which is important but not sufficient for a complete understanding of this ecosystem. Thus, measurements of large spatial- and temporal-scale variations of vegetation production, disturbances and their interaction are crucial to fulfill the existing data gaps. This is particularly important to understanding the role of this crucial ecosystem in the global carbon budget.

Remote sensing of vegetation production and disturbances is a critical measurement needed to extend the field level understanding of ecological, hydrological and biogeochemical processes to broader spatial and temporal scales in terrestrial ecosystems (Asner, 2004) and the different scales of energy, CO2 and mass exchange between ecosystems and atmosphere

Remote Sensing of Biomass in

White, 1983).

**2.2 Floristic composition and structure** 

*al*., 1996; Desanker *et al.*, 1997) dominate the ground-layer.

(Cauldwell & Zieger, 2000; Chidumayo, 1997).

the Miombo Woodlands of Southern Africa: Opportunities and Limitations for Research 79

soils is 0.3-3.8% (Chidumayo, 1989; Ribeiro & Matos*, unpubl. Data;* Sitoe *et al*., *unpubl. Data;* Walker & Desanker, 2004). Miombo soils have low concentration of organic matter with an average in the topsoil of 1% and 2% for dry and wet miombo, respectively (Chidumayo, 1997). This is a consequence of the abundant termite activities and frequent fire incidence

Fig. 1. Map of African vegetation, showing the miombo woodlands in dark green (Source:

Miombo have an estimated diversity of 8,500 species of higher plants, over 54% of which are endemic and 4% are tree species. Zambia is considered to be the centre of endemism for *Brachystegia* and has the highest diversity of tree species (Rodgers *et al.*, 1996). The diversity of canopy tree species is, however, low and characterized by the overwhelming dominance of trees in the genera *Brachystegia* (*miombo* in Swahili), *Julbernardia* and/or *Isoberlinia*  (Campbell *et al.*, 1996). Other important tree species in miombo include *Pseudolachnostylis maprouneifolia*, *Burkea africana*, *Diplorhynchus condilocarpon* among others. In mature miombo these species comprise an upper canopy layer made of 10-20 m high trees and a scattered layer of sub-canopy trees. The understorey is discontinuous and composed of broadleaved shrubs such as *Eriosema, Sphenostylis, Kotschya, Dolichos* and *Indigofera* and suppressed saplings of canopy trees. A sparse but continuous herbaceous layer of grasses, forbs and sedges composed of *Hyparrhenia, Andropogon, Loudetia, Digitaria* and *Eragrostis* (Campbell *et* 

Species composition and structure of miombo vary along the rainfall gradient across the region. In the dry miombo, canopy height is less than 15 m and the vegetation is floristically

(Carlson & Ripley, 1997; Goward *et al.*, 1985; Justice *et al.*, 1998; Running *et al.*, 1995; Schlesinger, 1996; Tucker *et al.* 1985). In areas where detailed and sufficient field data is scarce, as in much of the miombo context, the need for remote sensing data and techniques is even more important (Justice & Dowty, 1994; Malingreu & Gregoire 1996). Interpretation of the spaceborne data on land carbon stocks is needed, not only from the scientic point of view, but also within practical carbon management options mentioned in UNFCCC (Kyoto Protocol, REDD and REDD+).

To accurately measure objects on earth from the space several issues have to be considered including, the type and characteristics of remote sensing system, the spectral characteristics of the target objects, interactions between objects on earth, the statistical methods, among others. The advance of new generation of remote sensing such as IKONOS and QUICKBIRD optical sensors and, LiDAR and ALOS/PALSAR microwave sensors with high spatial resolution opens a new opportunity to improve understanding of miombo dynamics. These sensors, allow individual trees to be recognized and thus, large-scale biomass estimation in miombo woodlands. However, several constraints still exist and may limit the utilization of these data.

The aim of this chapter is to analyze the opportunities and constraints for the use of remote sensing techniques to estimate biomass (and carbon) in the miombo woodlands of southern Africa. The chapter also identifies research priorities for remote sensing of biomass in the miombo ecoregion.
