Added new FATES test that exercises the lightning dataset code

New test name:
ERS_D_Lm12.1x1_brazil.I2000Clm50FatesCruGs.cheyenne_intel.clm-FatesSuccIgnitions

I opted for a longer single-point test to allow several months of
lightning or ignition data to be read in.

NB. This test has uncovered a preexisting problem that I have documented
    in ExpectedTestFails.xml and which was missed before due to the
    short lengths of previous FATES tests.

cime_config/testdefs/ExpectedTestFails.xml

@@ -12,7 +12,7 @@
   </category>
   <category name="fates">
     <entry issue="NGEET/fates/#559" >FAIL SMS_Lm3_D_Mmpi-serial.1x1_brazil.I2000Clm50FatesCruGs.cheyenne_intel.clm-FatesHydro RUN</entry>
-    <entry issue="667"              >FAIL ERS_D_Ld5.1x1_brazil.I2000Clm50FatesCruGs.cheyenne_intel.clm-FatesHydro COMPARE_base_rest</entry>
+    <entry issue="991"              >FAIL ERS_D_Lm12.1x1_brazil.I2000Clm50FatesCruGs.cheyenne_intel.clm-FatesSuccIgnitions COMPARE_base_rest</entry>
     <entry issue="667"              >FAIL ERS_D_Ld5.1x1_brazil.I2000Clm50FatesCruGs.hobart_nag.clm-FatesHydro RUN</entry>
     <entry issue="667"              >FAIL ERS_D_Ld5.1x1_brazil.I2000Clm50FatesCruGs.izumi_nag.clm-FatesHydro RUN</entry>
     <entry issue="NGEET/fates/#510" >FAIL SMS_Lm3_D_Mmpi-serial.1x1_brazil.I2000Clm50FatesCruGs.hobart_nag.clm-FatesHydro MEMLEAK</entry>

cime_config/testdefs/testlist_clm.xml

@@ -2018,6 +2018,14 @@
       <option name="wallclock">00:40:00</option>
     </options>
   </test>
+  <test name="ERS_D_Lm12" grid="1x1_brazil" compset="I2000Clm50FatesCruGs" testmods="clm/FatesSuccIgnitions">
+    <machines>
+      <machine name="cheyenne" compiler="intel" category="fates"/>
+    </machines>
+    <options>
+      <option name="wallclock">00:40:00</option>
+    </options>
+  </test>
   <test name="SMS_Lm1" grid="f10_f10_musgs" compset="I1850Clm50BgcCropCmip6waccm" testmods="clm/basic">
     <machines>
       <machine name="cheyenne" compiler="gnu" category="aux_clm"/>

cime_config/testdefs/testmods_dirs/clm/FatesSuccIgnitions/README

@@ -0,0 +1,21 @@
+Testing the "successful ignitions" code, though reading the lightning
+dataset. The lightning and successful ignitions datasets use the same
+format and units, so handling is the same until you're in subroutine
+area_burnt_intensity where for "successful ignitions" the model selects
+currentSite%FDI = 1.0_r8
+cloud_to_ground_strikes = 1.0_r8
+while for lightning the model can use values < 1 for both.
+
+Lara Kueppers (UC-Berkeley) has access to a "successful ignitions"
+dataset for California that has not been placed in the repository of
+CTSM datasets as of 2020/6/6. That dataset can be used with the
+following settings:
+
+fire_method = 'spitfire'
+fates_spitfire_mode = 3
+stream_fldfilename_lightng = '.../data_UCB/observed/CA_monthly_ignition_number_1980-2016/ignition_1980_to_2016_monthly_20190801.nc'
+stream_year_first_lightng = 1980
+stream_year_last_lightng = 2016
+model_year_align_lightng = 1980
+
+where {...} = /fs/cgd/data0/slevis on izumi and /glade/work/slevis on cheyenne.

cime_config/testdefs/testmods_dirs/clm/FatesSuccIgnitions/user_nl_clm

@@ -0,0 +1,5 @@
+fire_method = 'spitfire'
+fates_spitfire_mode = 3
+stream_fldfilename_lightng = '/glade/p/cesmdata/cseg/inputdata/atm/datm7/NASA_LIS/clmforc.Li_2016_climo1995-2013.360x720.lnfm_Total_c160825.nc'
+hist_nhtfrq = 0
+hist_mfilt = 1