Construction of LAI and fAPAR maps of the Monarch Butterfly Biosphere Reserve and comparison with global maps
The Monarch Butterfly Biosphere Reserve (MBBR) is in a critical situation with levels of deforestation that threaten butterfly migration. Biophysical variables such as the leaf area index (LAI) and the fraction of photosynthetically active radiation absorbed by plants (fAPAR), can help diagnose the...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Online |
| Language: | spa |
| Published: |
Universidad Autónoma de Tamaulipas
2014
|
| Online Access: | https://revistaciencia.uat.edu.mx/index.php/CienciaUAT/article/view/293 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The Monarch Butterfly Biosphere Reserve (MBBR) is in a critical situation with levels of deforestation that threaten butterfly migration. Biophysical variables such as the leaf area index (LAI) and the fraction of photosynthetically active radiation absorbed by plants (fAPAR), can help diagnose the state of a forest. These variables can be measured in the field or, in case of a wide area, estimated from a linear regression fit between satellite images and field data, constructing high-resolution maps. There are also time series of low-resolution global maps of these variables available. In this work, we present the construction of LAI and fAPAR maps of the MBBR, using a SPOT image with spatial resolution of 20 m and field data. We compared them with GEOV1 maps of spatial resolution of 1 km, of the same variables and dates in order to assess the possibility of using them together. To build high-resolution maps, a field sampling was conducted in the summer of 2007, at 62 forest sites in the MBBR. Radiometric data of the corresponding pixel of a SPOT image of August 2007 were associated to each sample point; the areas with clouds were replaced by the same pixels in images taken on close dates. For each variable, 45 linear models were adjusted, from which the one with the highest concordance index was selected (0.45). These functions were applied to each pixel of the SPOT image to construct the maps, which were validated with field data, presenting acceptable relative errors (RMSEr, LAI: 20 %; fAPAR: 8 %). Comparison of these maps with GEOV1 maps of the same date, showed a significant correlation (r > 0.5) and an acceptable relative error (RMSEr, LAI: 19 %; fAPAR: 8 %). This result encourages the joint use of high resolution maps and time series. |
|---|
