terrestrial ecology website banner

Scoping Study Abstract for the Tree-Grass Project

Challenges and Opportunities in Remote Sensing of Global Savannas: A Scoping Study for a New TE Field Campaign

Niall Hanan, Colorado State University

Global savanna biomes are spatially and temporally complex systems in which woody
vegetation (trees, shrubs) and herbaceous vegetation (grasses, forbs) both contribute
significantly to system level functions such as primary production, carbon, water and
nutrient cycling. The tropical and temperate savannas represent ~30% of the global land
area and support unique communities of pastoral and agricultural peoples alongside wild
and domestic herbivores. Savannas are subject to highly dynamic interactions between
climate, soils, fire and herbivory, which depend strongly on human management and land
use, and will be significantly affected by future changes in climate. Savanna regions are
also subject to directional changes in the balance between woody and herbaceous cover
(shrub encroachment) that remain poorly understood and can have large impacts on land
surface-atmosphere interactions and system biogeochemistry. The production of the herbaceous layer, which naturally consists largely of grasses, has been co-opted by
humans in many regions to form vital centers for grain production.

In earlier years, NASA has invested in several research programs in savannas (HAPEX-
Sahel, SAFARI, dry end LBA). However, despite these activities, remote sensing of
mixed tree-grass systems remains challenging because of the separation of the vegetation
into two distinct layers (woody species of varying cover, density, height, leaf area and
biomass, over a herbaceous layer of varying density, cover and leaf area), and because
the woody and herbaceous layers can have distinct and contrasting seasonality,
physiology and phenology which vary in both space and time. Further, fire dynamics in
savannas require not only assessment of the timing, extent, and intensity of fires, but also
their impacts on vegetation structure. Without reliable methods to adequately assess
vegetation structure in savannas, parameterization of higher order models of vegetation
function (primary production, vegetation dynamics, water and energy balance) is difficult
and model results unreliable.

While there has been substantial research in savannas, the complexity of savanna
vegetation limits the effectiveness of model-data integration, scenario and predictive
analysis. Recent advances in active and passive remote sensing and, in particular, the
potential for combined Lidar and synthetic aperture radar (SAR) in the proposed
DESDynI mission, offer potential for significant advances in capturing the dynamics of
these systems and delivering accurate system level predictions into the decision-making
and policy domains. There is a need for a coordinated and targeted field and remote
sensing campaign in appropriate locations to integrate existing radar, lidar and passive
(broad-band and hyperspectral) remote sensing with modeling and field measurement to
realize this potential and prepare the way for new instruments planned as part of the
Decadal Survey missions. We propose a scoping exercise for a future field activity
focused on the challenges and opportunities for remote sensing in mixed tree-grass
systems. We will use input from NASA program managers and instrument teams,
savanna ecologists and the modeling community, to develop recommendations for
NASA, detailing optimal approaches and geographical locations for a remote sensing
oriented field program focused on the savanna biome.

Our proposal is relevant to NASA Terrestrial Ecology program goals for scoping studies
(subelement 2) and studies of vegetation structure using radar and lidar (subelement 1).
The study aims to fully realize the potential of satellite remote sensing in the
measurement and modeling of savanna vegetation structure and dynamics. It will
develop a program of research focused on new opportunities in active remote sensing and
how these can be utilized in synergy with new and historical passive remote sensing.