Using MODIS BRDF Products to Estimate LAI

1
Using MODIS BRDF Products to Estimate LAI

• Tang Shihao
• Department of Geography, Beijing Normal
  University
• National Satellite Meteorological Center , CMA

2
Content

• Introduction
• Principle and method
• LAI distribution of China
• Validation
• Discussion and conclusion

3
Introduction

• Inverse method of LAI
• VI based method
• Physical model based method
• Computer iteration
• Look-up table
• Hybrid method

4
Principle and method

• LAI can be expressed as the multiplication of
  leaf area density(µ) and canopy height(z)
• LAIµZ
• The analytical solution of folidge density is
  given by Miller (1967) as

5
Principle and method

• The expression of LAI can be derived from µ and z

6
Principle and method

• for multiple angle measurements, the discrete
  approximation of the above equation is applied

7
Principle and method
8
LAI inversion process based on MODIS BRDF/ALBEDO
product
9
(No Transcript)
10
Determination of saturation and background NDVI
11
Determination of saturation and background NDVI
12
Integration angle setting

• Table 1 shows that the average integration step
  applied in LAI2000 is 15?,the maximal integration
  angle is 68?, and the integration step of the
  maximal integration angle is equivalent to
  30?.The same integration steps and angle
  configuration can be used in our algorithm for
  the consideration of speed. In this paper, we
  adopt 5? average integration step, 75? maximal
  integration angle and 15? integration step of the
  maximal integration angle in order to get good
  performance.

13
LAI distribution in China
This algorithm
MODIS
14
validation
15
validation
16
(No Transcript)
17
discussion
18
conclusion

• An important significance of this work lies in
  providing a new train of thoughts or frame that
  is different from traditional methods. Under this
  frame, many modules in it can be further studied
  and improved, for example, the BRDF algorithm,
  the gap fraction algorithm etc.. Now, a new
  practical and parameter independent gap fraction
  algorithm is being developed, and we hope that it
  can greatly improve the performance of our
  method. Although the current algorithm is
  unlikely better than traditional methods because
  of the restriction of some products precision,
  such as BRDF product, we can still foresee that,
  with the improvement of remotes sensing product
  precision and with the further study of the
  algorithm itself, its performance will be better
  and better.