This page lists known issues for HadEX3, HadEX2 and GHCNDEX.
GHCNDEX
Due to the nature of spatial interpolation the CDD and CWD indices can have temporal inconsistencies from one year to the next, particularly in dry regions. Users should interpret these indices very cautiously in dry regions (e.g. in the middle latitude dry zones).
HadEX3
Australian temperature indices contain data from GHCND stations in addition to ACORN-SAT stations. This is different from HadEX2 and therefore the gridded output is not directly comparable.
Precision problems with station coordinates in Slovenia (originating from ECA&D indices template files). This may have a minor impact on the gridded output over this region.
Station names in HadEX3 are typically the station ID's from the host dataset that HadEX3 draws from and hence identifying stations by their name (e.g. "Brisbane Airport") is often not possible and station ID's must be used.
PRCPTOT for GHCNDEX is only calculated annually but for HadEX3 is calculated monthly. Since HadEX3 uses GHCNDEX indices over Australia the monthly gridded PRCPTOT values in HadEX3 do not include any data from GHCNDEX, even though annual gridded PRCPTOT values in HadEX3 do include data from GHCNDEX.
Due to the nature of spatial interpolation the CDD and CWD indices can have temporal inconsistencies from one year to the next, particularly in dry regions. It is also possible that some years are shown as missing even when sufficient data exists. This is because in a very few cases consecutive dry days can span multiple years, in which case the total number of dry days is reported in the last year of the dry spell and all preceding years in the dry spell are marked as missing. This avoids double counting but can cause confusion in regions or datasets with missing data. Users should interpret CDD very cautiously in dry regions (e.g. in the middle latitude dry zones).
HadEX2
Some station longitudes are incorrect (e.g. UK, western France, Portugal, Brazil) due to incorrect conversion of minutes/seconds to decimal degrees. This leads to a margin of error of less than one degree of longitude but in some cases results in, for example, stations being located over water. The effect of this is under investigation.
Incorrect angle calculation in Angular Distance Weighting (ADW) method which may have a minor impact on the gridded output.
Precision problems with station coordinates in Slovenia (originating from ECA&D indices template files). This may have a minor impact on the gridded output over this region.
Due to the nature of spatial interpolation the CDD and CWD indices can have temporal inconsistencies from one year to the next, particularly in dry regions. It is also possible that some years are shown as missing even when sufficient data exists. This is because in a very few cases consecutive dry days can span multiple years, in which case the total number of dry days is reported in the last year of the dry spell and all preceding years in the dry spell are marked as missing. This avoids double counting but can cause confusion in regions or datasets with missing data. Users should interpret CDD very cautiously in dry regions (e.g. in the middle latitude dry zones).