Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • classes
  • ctdas-cosmo
  • dynamicFFmodel
  • feature-STILT
  • griddedstatevector
  • ivar
  • karolina
  • liesbeth
  • master default protected
  • near-real-time
  • old-master
  • package
  • py-3
  • remotec
  • trunk
  • wouter
16 results
Created with Raphaël 2.2.09Dec11May29Apr1215Mar718Feb1714Jan17Nov21Oct25Sep2423218431Aug2420191815141328Jul222095125Jun24527May262221201915Apr10Mar26Feb22Jan713Oct25Jul1416May4Apr2Mar126Feb1931Jan30292819Dec16Oct27Aug2322212029Jul171615121030Jun28272625242017131211107621May201387215Apr98428Mar262522211927Feb12129Nov201531Oct3029252126Sep4331Aug302823106231Jul27262524232013121020Jun121131May29252421181716151411926Apr25Added rc file for CTE2016, including option to use 2 ObsPack products.revised the way for calculating BC footprintsUpdates in analysis for producing graphs etc. for CTE2015 website, plus some minor changes for Eric's SAM runs.Added mdm scaling with number of observations per day (except for aircraft)Added copying of *nc files from da/analysis to cloned runFaster method to look up obs_nums, especially useful for GLOBALVIEWplus ObsPack hourly time seriesforward simulation module of the CT-COS system (COS flux optimization)BC sensitivity with 4 parametersupdatedBC footprints normalized caseadd more information to the sampling outputsminor revision for the forward simulation modulefixed a bug for EMP case (problem on adding BC parameter)hysplit.rc filetime-involved BC sensitivity for HYSPLIT-NAM12time-involved BC sensitivity for WRF-STILTerror occurs when an observation is taken exactly at midnight UTC (00:00). This causes a problem to calculate the n value because hr and mi variables are empty. I made a quick fix by setting n value to 24 if an observation is taken at midnight.latest revisions from ivar. fixed scalefac alignment and added log statementsrevisedfixed morerevised errors on time mismatch between fluxes and footprintsfixed problem on flux, bc parameter, and footprint backward mismatchbug fixed for BC parameter updatingbug fix BC update parametersbug fix in residual calculation in serial_least_squaresbug fix in serial function. KG values were only set zero for the last lag instead of both lagsRemoved bug in add_model_data_mismatch. Without the do-not-simulate flag in sites_and_weights.rc file the complete data list was returned instead the trimmed down list that is filtered.Removed bug in add_model_data_mismatch. Without the do-not-simulate flag in sites_and_weights.rc file the complete data list was returned instead the trimmed down list that is filtered.summarized tableto involve 2011 January or more footprintsremoved argument from add_model_data_mismatchco_filter keysAdded a CO filter to exclude observations with high levels of carbon monoxideAdded co.filter argument to the function add_model_data_mismatchbug in serial_least_squares. self.KG[-1]==0. should become self.KG[-1]=0.set default nprocs to 2.removed '-1' factor to calculate BC. BC parameter is already relative to 0.0Removed test covariance. Defined a default BC uncertainty of 2 ppm^2 in the covariance matrix. Made sure deviations in BC parameter are now relative to 0.0added carbontracker_griddedNA in stilt as a referenceexpand_fluxes and expand_molefractions do now come out of the da/stilt/ directory