steps for data downloads

How to download files: Step-by-Step

These methods may not  be the only way! See the source for more information. Other sites may also have the data (NCAR, for example) and those methods would be different than those below. 

MERRA-2 (Using ftp subsetter)

1. ​Go to page http://disc.sci.gsfc.nasa.gov/daac-bin/FTPSubset2.pl
2. From the list, find type of data you are looking for. 3D, daily, monthly, etc. Note assimilation is term used for what many researchers know as analysis.
3. Optionally select a region to subset.
4. Select​ a start and end month. You cannot get all years so select a small subset to start with.
5. Select a variable(s)
6. Optional- select a remapping type and grid to specify grid size.  Some match other datasets. Some may not be obvious and you may need to try them and see what you get.
7. Select an output format. netcdf4-classic can be read by many applications.
8. Select "start search". You will get a list of files. You can save this list and run wget --content-disposition -i wget_n3LWiFa9
9. You can also download a file and look at it to see if it's what you need.
10. To get all files, you can run a script to get each file with a wget -content-disposition command and fill in the dates you need. Or you can use an editor to generate the list, changing the filename dates.
11. Use name/password: You need to use a name and password to get the data. wget --user username --password user password
12. Check to see that the files are downloaded completely.

​A few notes:

• There are different streams so you have to change those depending on the year.
• You can change the output file names to whatever you want. 
• You must close the window of the subsetted data search page in order to perform a new search!

JRA-55:

This out out of date. NCAR has the data there. 

1. Register for dataset. You will get a user name and password. http://jra.kishou.go.jp/JRA-55/index_en.html#usage
2. Get the data via ftp with your name/password.  See http://jra.kishou.go.jp/JRA-55/index_en.html#download
3. From the command line, ncftpls -u name -p password ftp://ds.data.jma.go.jp/JRA-55/Hist/Monthly/anl_p125 for example. Or use ncftpget
4. Data is divided into directories: http://jra.kishou.go.jp/JRA-55/document/JRA-55_handbook_LL125_en.pdf

ERA-Interim:

1. Go to this webpage: http://apps.ecmwf.int/datasets/data/interim-full-daily/levtype=sfc/
2. Login before getting data. You will need to apply for a login. Safari may have issues with ERA pages on MACS; Firefox will work.
3. On the left column of the webpage above, select the type of analysis you need (time scale, data type).
4. Select a month, variable. View MARS request. You will see a python script sample. That script will list the parameters you need to download the data files from a script. The parameters can be used by python or Perl scripts or similar languages.
5. Download required modules depending on what language you are using. See this webpage https://software.ecmwf.int/wiki/display/WEBAPI/Access+ECMWF+Public+Datasets
6. Edit your script to download the complete set of dates/variables you need. Include the required modules in your script. 

20CRV2c:

1. ftp to ftp.cdc.noaa.gov
2. cd /Datasets/20CRV2c/
3. get files.  ncftp or similar command line client is preferable as the whole file is more likely to download than using a browser.  Data is divided into directories by time scales and type.
4. For the browser link, use https://downloads.psl.noaa.gov/Datasets/20thC_ReanV2c/

20CRV3:

1. ftp to ftp.cdc.noaa.gov
2. cd /Datasets/20CRV3/
3. get files.  ncftp or similar command line client is preferable as the whole file is more likely to download than using a browser.  Data is divided into directories by time scales and type.
4. For the browser link, use https://downloads.psl.noaa.gov/Datasets/20thC_ReanV3/