I am working on making LAIv input files for MEGAN 3.2. I am familiar with LAI but not so much with LAIv and MEGAN. I understand that LAIv is the ratio of LAI to vegetation cover fraction (VCF). My question is if I want to produce a LAIv file for January, should I use January VCF data or the maximum VCF data from the whole year? Because I have noticed that if I use January values for both LAI and VCF, the fact that the VCF is smaller in the winter months results in high LAIv values in some areas in January, even higher than the summer LAIv values. I’m concerned that this will make MEGAN to overestimate winter BVOC. In fact, I still don’t understand why LAIv is used for MEGAN instead of LAI, even though I’ve read a few papers on the subject. In my opinion, LAI is more representative to quantify the amount of foliage in each grid than LAIv, which is influenced by VCF. If anyone can make any suggestions or correct my mistakes, I would appreciate it!
That’s interesting that you’re seeing places where LAIv is higher in winter. I don’t see this over North America for our usual configuration, presumably because LAI decreases along with veg fraction. Winter LAIv minus Summer LAIv:
I’m not positive why MEGAN is formulated for LAIv instead of LAI, but you can imagine a grid cell where only half of the cell is covered in vegetation. It would make sense to simulate process for that part of the grid cell with dense canopy and then flux the emissions into the grid cell, instead of assuming a less dense canopy over the entire cell.
Can I ask what data you’re using that’s showing higher LAIv in winter than summer?
Your image does show that the LAIv in winter is always smaller than in summer. But my results in the East Asian are quite different from that. I don’t know if it’s because I’m using the wrong dataset or my processing is wrong.
I calculated LAIv for 2017. The LAI I used was from GLASS LAI (version: V60, MODIS_0.05°) and the VCF I used was from GLASS VCF (version: V40, MODIS_0.05°), the web link is https://www.glass.hku.hk /download.html. I chose to use these datasets because they are currently popular datasets that are widely downloaded, used and validated. The way I processed it was: 1. exclude grids with VCF below 0.1 in ARCGIS PRO software because too low VCF can lead to unreasonably high LAIv values (I’m not sure if it makes sense for me to do this, maybe I shouldn’t exclude any grids or should exclude more grids, e.g., exclude grids with VCF below 0.3?)
I scrutinized some of the grids and found that the low VCF is the cause of the LAIv being larger in some areas in winter than in summer. For example, there is a grid where the LAI and VCF in summer are 4 and 0.8 respectively, so LAIv is calculated as 4/0.8 = 5. In winter, the same grid has LAI and VCF of 1 and 0.15 respectively, so LAIv is calculated as 1/0.15 = 6.66 >5. The situation is even more exaggerated for some other grids. I have attached the images below:
Thank you for your patience. I saw this on the MEGAN official page:“Note that these LAI data must first be converted to LAIv (LAI divided by the vegetation cover fraction) before used in MEGAN”. Does the term “vegetation cover fraction” actually refer to FPAR and not VCF?
Hi Jeff,
There are several global products for the VCF, such as 1 km MODIS-based Maximum Green Vegetation Fraction product from USGS, which can be used for the LAIv calculation. It is also quick to do the calculation and then compare with the LAIv for the North America dataset provided by the MEGAN group (which would pinpoint the relative error and bias in the calculation).
