USE OF LIDAR DATA

1
USE OF LIDAR DATA TO ASSESS CANOPY STRUCTURE
Eugenia González ECL 290 Fall 2004
2

• OUTLINE
• Use of lidar data to characterize canopy
  structure
• Description of the technique
• Potential applications
• RS studies in Tropical Forests
• Example La Selva study
• Some caveats

3
Lidar (light detection and ranging)

• Active remote sensing technique
• Uses pulses of laser light (900-1064 nm)
• Measures the round-trip for a pulse to travel
  between the sensor and the target
• Provides a range (distance) between the
  instrument and the object

4
Conceptual basis of operation
5
Two types 1. Discrete-return lidar 2.
Waveform-recording
6
Applications for vegetation studies

• Provide data on the vertical distribution of
  intercepted surfaces ? Aboveground biomass
  estimates ? Carbon studies
• Spatial pattern of canopy height and cover
• Derived metrics stem diameter, basal area, mean
  DBH
• Metrics from lidar sensitive to different
  land-use histories

7
Two general problems

• Exact elevation of ground surface may be
  difficult to estimate if understory is dense
  enough to occlude ground surface
• Height of the canopy may be underestimated if
  the top portion of the crowns do not have
  sufficient area to register a return signal

8
Some Priority Research Goals for RS in
Tropical Forests

• Identification and characterization of secondary
  forests
• Documentation of the rate and extent of
  deforestation

9
Why are RS studies so difficult in the tropics?

• Frequent overcast conditions
• Different stages of regrowth are hard to
  discriminate

• Dry tropics deciduousness vary with climatic
  conditions and successional status
• Rapid saturation of common indices

10
La Selva Biological Station study

• Northeastern Costa Rica, 30-150 m elevation
• Tropical Wet Forest
• Mixture of forest plantations, secondary and
  old-growth patches

11
La Selva Biological Station study
OBJECTIVE Assess the potential of lidar data to
initialize a height-structured ecosystem model
(ED) to estimate carbon stocks and fluxes
METHODS Lidar data (LVIS) collected at 8 km of
altitud, footprint size of 25 m diameter. Total
swath width of 1 km
12
METHODS
13
RESULTS
Distribution of mean canopy heights measured by
LVIS from areas with different land-use histories
14
RESULTS (cont)
Good agreement between field, regression and ED
estimates
up to a mean canopy height of 29 m
15
RESULTS (cont)
Estimates of aboveground Carbon stocks
16
RESULTS (cont)
Estimates of mean aboveground Carbon flux
17
From a related study
Detailed landscape pattern in ABG biomass
18
CONCLUSIONS

• Lidar data of canopy height was effective to
  initialized an ED model
• Better estimates for Carbon stocks and fluxes
  since the structure represents actual
  successional state of forest, not only potential
  vegetation
• Important constraint on model estimates

19
References

• Drake JB, RO Dubayah, DB Clark, RG Knox, JB
  Blair, MA Hofton, RL Chazdon, JF Weishampel, SD
  Prince. 2002. Estimation of tropical forest
  structural characteristics using large-footprint
  lidar. Remote Sensing of Environment 79 305-319.
• Hurtt GC, RO Dubayah, J Drake, PR Moorcroft, SW
  Pacala, JB Blair, MG Fearon. 2004. Beyond
  pontential vegetation Combining lidar data and a
  height structured model for Carbon studies.
  Ecological Applications 14 873-883.
• Lefsky MA, WB Cohen, GG Parker, DJ Harding. 2002.
  Lidar remote sensing for ecosystem studies.
  BioScience 52 19-30.
• Mather PM. 2004. Computer Processing of
  Remotely-Sensed Images. An introduction. 3rd
  edition. John Wiley and Sons. West Sussex,
  England.
• Sánchez-Azofeifa GA, KL Castro, B Rivard, MR
  Kalascka. 2003. Remote sensing priorities in
  tropical dry forest environments. Biotropica 35
  134-142.

20
Thank you!