Identify the data you want
- SFC, 6 parameters
- Time: 00, March 2017
- all steps (0/to/90/by/1, 93/to/144/by/3 and 150/to/360/by/6)
- all numbers (1/to/50)
- Area: Europe
- Output grid: regular lat/lon 0.5/0.5
Use MARS list to find out the size
Using the MARS catalogue and the "View MARS request" functionality you can create a list request of all the data you want which would tell you the size and distribution of the raw data in the archive. Note the LIST verb and the OUTPUT = COST keyword.
LIST,
CLASS = OD,
TYPE = PF,
STREAM = ENFO,
EXPVER = 0001,
LEVTYPE = SFC,
PARAM = 134.128/151.128/165.128/166.128/246.228/247.228,
DATE = 20170301/to/20170331,
TIME = 0000,
STEP = 0/1/2/3/4/5/6/7/8/9/10/11/12/13/14/15/16/17/18/19/20/21/22/23/24/25/26/27/28/29/30/31/32/33/34/35/36/37/38/39/40/41/42/43/44/45/46/47/48/49/50/51/52/53/54/55/56/57/58/59/60/61/62/63/64/65/66/67/68/69/70/71/72/73/74/75/76/77/78/79/80/81/82/83/84/85/86/87/88/89/90/93/96/99/102/105/108/111/114/117/120/123/126/129/132/135/138/141/144/150/156/162/168/174/180/186/192/198/204/210/216/222/228/234/240/246/252/258/264/270/276/282/288/294/300/306/312/318/324/330/336/342/348/354/360,
NUMBER = 1/to/50,
OUTPUT = cost,
TARGET = list.txt
When we run the list action this is the output:
size=4484574297000; # 4.48 TB number_of_fields=1348500; online_size=799241928660; off_line_size=3685332368340; number_of_tape_files=93; number_of_disk_files=281; number_of_online_fields=240330; number_of_offline_fields=1108170; number_of_tapes=32;
The data is too large: 4.48 TB and split across 32 tapes.
Use the MARS catalogue to know the data distribution in files and tapes
Using the MARS catalogue we browse the data we want to get until we get to the final stage which gives us a selection and several options: http://apps.ecmwf.int/mars-catalogue/?stream=enfo&levtype=sfc&time=00%3A00%3A00&expver=1&month=mar&year=2017&date=2017-03-01&type=pf&class=od
For this particular case we can choose different "Parameter", "Number" and "Step". This means that all the fields you can choose in this page are in the same tape file. Therefore users should get as much data as possible from this page.
If we run the list of the data that we need and appears in the final stage page we get:
LIST,
CLASS = OD,
TYPE = PF,
STREAM = ENFO,
EXPVER = 0001,
LEVTYPE = SFC,
PARAM = 134.128/151.128/165.128/166.128/246.228/247.228,
DATE = 20170301,
TIME = 0000,
STEP = 0/1/2/3/4/5/6/7/8/9/10/11/12/13/14/15/16/17/18/19/20/21/22/23/24/25/26/27/28/29/30/31/32/33/34/35/36/37/38/39/40/41/42/43/44/45/46/47/48/49/50/51/52/53/54/55/56/57/58/59/60/61/62/63/64/65/66/67/68/69/70/71/72/73/74/75/76/77/78/79/80/81/82/83/84/85/86/87/88/89/90/93/96/99/102/105/108/111/114/117/120/123/126/129/132/135/138/141/144/150/156/162/168/174/180/186/192/198/204/210/216/222/228/234/240/246/252/258/264/270/276/282/288/294/300/306/312/318/324/330/336/342/348/354/360,
NUMBER = 1/to/50,
OUTPUT = cost,
TARGET = list2.txt
size=144663687000; number_of_fields=43500; online_size=43232826000; off_line_size=101430861000; number_of_tape_files=3; number_of_disk_files=10; number_of_online_fields=13000; number_of_offline_fields=30500; number_of_tapes=3;
In this case the raw data (without post-processing) is ~144GB, which would be reduced if you interpolate to lat/lon and/or filter the area.
The data is split in 3 different tapes, which is sensible.
To know the size of the file that will be transferred to the system, you can download 1 field using the Post-processing keywords (usually GRID and AREA). Then you should multiply the size of the 1 field file by the number of fields that you want to retrieve in 1 request: number_of_fields=43500.
Now that we know the size of the data that we want to retrive in one go from the same hypercube we can start the study of the way to iterate to get the full month.
Split the request in sensible chunks iterating through the correct keywords
Using the browser we know how the data is distributed in the MARS tree. Now we focus on the current selection section. In this case:
From the top to the bottom we have to start iterating from the inner loop to the outer loop "Time", "date", ...
This is an example BASH script to loop 1 time/ 1 day in one go for one month
#!/bin/bash
#this example will filter the area of Europe (N/W/S/E) and interpolate the final fields to a lat/lon 0.5/0.5 degrees
AREA="73.5/-27/33/45"
GRID="0.5/0.5"
PARAMS="134.128/151.128/165.128/166.128/246.228/247.228"
TIMES="0000"
YEAR="2017"
MONTH="03"
#date loop
for y in ${YEAR}; do
for m in ${MONTH}; do
#get the number of days for this particular month/year
days_per_month=$(cal ${m} ${y} | awk 'NF {DAYS = $NF}; END {print DAYS}')
for my_date in $(seq -w 1 ${days_per_month}); do
my_date=${YEAR}${MONTH}${my_date}
#time lop
for my_time in ${TIMES}; do
cat << EOF > my_request_${my_date}_${my_time}.mars
REQUEST,
CLASS = OD,
TYPE = PF,
STREAM = ENFO,
EXPVER = 0001,
LEVTYPE = SFC,
GRID = ${GRID},
AREA = ${AREA},
PARAM = ${PARAMS},
DATE = ${my_date},
TIME = ${my_time},
STEP = 0/1/2/3/4/5/6/7/8/9/10/11/12/13/14/15/16/17/18/19/20/21/22/23/24/25/26/27/28/29/30/31/32/33/34/35/36/37/38/39/40/41/42/43/44/45/46/47/48/49/50/51/52/53/54/55/56/57/58/59/60/61/62/63/64/65/66/67/68/69/70/71/72/73/74/75/76/77/78/79/80/81/82/83/84/85/86/87/88/89/90/93/96/99/102/105/108/111/114/117/120/123/126/129/132/135/138/141/144/150/156/162/168/174/180/186/192/198/204/210/216/222/228/234/240/246/252/258/264/270/276/282/288/294/300/306/312/318/324/330/336/342/348/354/360,
NUMBER = 1/to/50,
TARGET = enfo_pf_${my_date}_${my_time}.grib
EOF
mars my_request_${my_date}_${my_time}.mars
if [ $? -eq 0 ]; then
rm -f my_request_${my_date}_${my_time}.mars
fi
done
done
done
done