Reduce MODIS L1b/L2 test case size for better test performance

[djhoese]

As described in #1883 the MODIS tests use a lot of memory and take a long time to run. This isn't generally a problem, but for lower resource systems they may actually be unable run the tests. This PR attempts to fix this by reducing the unnecessarily large number of scans used in the test data to only be 3 scans worth.

CC @avalentino. I doubt you can test this easily, but this is a good enough change that it should be merged on its own.

• Closes Test failure on i386 and armhf #1883
• Tests added

[codecov]

Codecov Report

Merging #1966 (a7cec8a) into main (279631e) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1966   +/-   ##
=======================================
  Coverage   93.53%   93.53%           
=======================================
  Files         279      279           
  Lines       41312    41312           
=======================================
  Hits        38642    38642           
  Misses       2670     2670           

Flag	Coverage Δ
behaviourtests	4.77% <0.00%> (ø)
unittests	94.07% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
satpy/tests/reader_tests/_modis_fixtures.py	97.50% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 279631e...a7cec8a. Read the comment docs.

[coveralls]

Coverage remained the same at 93.999% when pulling a7cec8a on djhoese:test-smaller-modis into 279631e on pytroll:main.

[djhoese]

I forgot to say this makes it so the modis_l1b tests take ~3s instead of ~55s.

[avalentino]

Hi @djhoese i confirm that the patch works. Thanks a lot.

[mraspaud]

LGTM