Hello there.
Im trying to run the preprocessor tool for Megan 2.1.
It is my understanding that I need to run both the leaf index area and plant functional type pre processors to regrid the input data to my domain of interest.
Here is the thing: My domain is in South America and in the Megan website there are only datasets for North America.
Does anyone know where can I find these datasets for other regions? Or if they are not essential?
Thank you
Sol
Hi,
For the leaf area index, you can use a MODIS product containing the LAI. You have to regrid the data to your domain.
About the plant functional type, you can use dataset from MEGAN3.1 datasets.
I do recommend to use MEGAN3.1 since its datasets (except the leaf area index which you can retrieve from MODIS) have global resolution.
Hello there,
Thank you for your answer. I have a few questions:
So…if I get my leaf area index data from Modis…can I then use the Megan preprocessors to regrid it or should I find another way?
Regarding the plant functional type…for MEGAN3.1 I dont see datasets with the exact same name, but “Growth form and ecotypes”…is this the same? Or am I missing something? Can they be used with MEGAN2.1?
Thank you so much!
You can use megan preprosesor for regriding. However, their program works in a specific manner and reads variable with specific name. So, you have to mosaic all MODIS HDF files, and them translate them into correct NETCDF format which is readable by MEGAN preprocessor. Alternative method is to skip all steps in between and create your own LAI NETCDF file following the sample file in the MEGAN model.
The approach in MEGAN3.1 is different than MEGAN2.1. I have no idea about this question.
I understand.
Thank you very much!
I would like to add on the differences between MEGANv3.1 (actually now it is MEGANv3.21) and MEGANv2.1. The parameterization of calculating BVOC differs between these two versions (You can search for MEGAN3 project report in Google if you would like to know more).
In addition, for LAI and PFT over South America, you can use MODIS products to create these nc files by yourself (Please note that there are a few steps that you have to figure out by yourself including mosaic and convert them to netcdf format). Also, please note that LAI used in MEGANv2.1 is actually LAIv (which means you have to include vegetation coverage). For more details, please refer to MEGAN official website on Professor Guenther’s website.
Hope it helps.
Ryan
Thank you so much for your comments.
Regarding your comment about getting LAI and PFT data over South America from MODIS products…I found MODIS datasets that have leaf area index, but Im having trouble finding vegetation coverage (To compute LAIv…I found “greeness_fraction”, but I dont think is exactly the same?) and PFT data. Maybe they are under different names? Also, the dataset for LAI that I found does not have the same horizontal resolution that the ones provided with Megan…is this a problem?
I considered, for simplicity, running MEGANv.3.1 instead (As they have global Growth Form and Ecotypes datasets so I would only need to use MODIS datasets for the LAI)…But I found that Growth Form and Ecotype datasets are not available, despite being mentioned in the website (Either they dont exist or I need permission to access them which I asked but have not been granted). Do you know if there is another place where this datasets are hosted or from where can I download them?
Again, thank you very much
Sol
For the missing files that you mentioned, these files are not available due to the limit of Google Drive on our UCI email address. Please DM me (wuk15@uci.edu) your email address and let me know which files you can’t find, and then I can share these files with you.
Best,
Ryan
The same resolution is not a requirement. You can ncdump the NC files over North America then you will see the resolution is not consistent with MEGAN, because the pre-processor is a regrid executable.
Ryan, thank you very much. I just emailed you detailing which files Im missing,
Again, thank you!
Dear SoIL, I was wondering if you managed to run MEGAN2.1 in South America. I have spent a lot of time trying to do it for my doctoral thesis, but MEGAN can’t read my input files. I think it might be because the names are different, but I’m not sure. If you succeeded, could you please tell me the names you used for each variable in the EF, LAI, and PFT files?
Thank you very much for your help.
Hello, Valeria.
I gave up with MEGAN2.1 and instead end up using MEGAN3.2. Is much easier this way as the inputs needed are:
Leaf Index Area: This files can be made for South America with MODIS data as discussed here, but with this MEGAN version they are not mandatory as you can choose to read them from the MCIP files (and that is what I did).
Growth form and ecotypes: There are available global files for this inputs. The last time I checked they werent available in the web page, but someone (I cant remember who) kindly provided me with the following link where they are uploaded:megan_inputs.tar.gz - Google Drive)
If you can, maybe you can consider switching up MEGAN versions as I did
You can try “ncdump -h filename.nc” to double-check the variable name in your input file. Make sure your file consistent with MEGAN sample files (everything should be consistent).