Disclaimer for climate data

Please be advised that certain irregularities have been observed regarding two parameters in the climate data service. The two parameters are 

  • potential evaporation calculated with the Penman equation (pot_evaporation_penman), and
  • the Drought Index (drought_index).

DMI therefore recommends that instead of using the parameter potential evaporation calculated with the Penman equation, please use the parameter potential evaporation calculated with the Makkink equation (pot_evaporation_makkink) instead. 


When the parameter Drought Index is calculated, potential evaporation calculated with the Penman is used as input in the equation for the Drought Index model. This means that there may be a tendency toward a higher Drought Index during the summer.




Table of contents:






About Spatial Resolutions

The climateData service includes data in different spatial resolutions for Denmark.

Spatial resolutionDescription

10kmGridValue  

climate data in a 10 x 10km grid resolution covering danish land areas
20kmGridValue

climate data in a 20 x 20km grid resolution covering danish land areas

municipalityValueclimate data for the Danish municipalities
countryValueclimate data for Denmark


The data is based on station data from Danish stations, which  is interpolated into a 1x1km grid net covering all of Denmark’s land area using a modified inverse-distance algorithm. The grid is based on Det Danske Kvadratnet's 1x1km grid with the grid name DKN_1km_ETRS89. The calculated values in each grid cell depends on the values of the nearest surrounding station found in 8 sectors. One value is calculated for each area (e.g. one municipality) from the grid cells covering the jurisdiction area combined with a weighting of how much of each grid cell is within the area. The stations are weighted in relation to the distance and climatological comparability (distance to sea) to the grid cell. Aggregated grid cells are used to compile datasets of different spatial resolution.

The different spacial resolutions are calculated/derived from stationValue data as it becomes available. The data is recalculated when the quality controlled stationValue data has been quality controlled by our climatologists. In some instances area data is based on additional parameters, which are not quality controlled. Please see the parameter list for an overview over which parameters are available and whether or not they are quality controlled by our climatologists.

Spatial resolutions are not available for Greenland, as an interpolation algorithm cannot be applied due to the complicated circumstances of the large geographical area, complex topography and low station density.


Interpolation Algorithm – Grid Data

The stationValue dataset is used to create a 1x1 km grid net covering all of Denmark’s land area. The grid net is calculated from a modified inverse-distance algorithm and is based on all stations that DMI has at its disposal, including both DMI and third party stations. In order to understand the algorithm one must consider the local Danish climate. The overall factor that generally spoken has the largest influence on the local climate in Denmark is the distance to the sea. The uneven station coverage of Denmark's area represents a challenge if using a classic non-weighted interpolation. Areas with bad coverage are in risk of being affected by stations located in an area that climatologically seen is not representative for the point of interpolation. Therefore a modified inverse-distance algorithm is applied, where the value in each grid cell depends on the values of the nearest surrounding stations found in 8 sectors. The stations are weighted in relation to the distance and climatological comparability (distance to sea) to the grid. 

The coast-/inland climate distribution follows figure 1 below:


Station Selection

Each grid cell value is calculated from weighted station values. A grid cell is not represented by every station equally. Station selection is needed since some stations may be located in a different climate zone or add undesired effects to the grid cell value. The algorithm selects the nearest stations in 8 sectors. 

In the following, the station selection method is illustrated for the case shown in Figure 2. The green triangles illustrate stations and the red square is the grid cell to be calculated. The algorithm creates four lines passing through the center of the grid cell, creating eight sectors (ENE, NNE, NNW, WNW, WSW, SSW, SSE, ESE) and finds the nearest station in each sector - aquamarine triangles in Figure 2 right panel.

The calculated values in each grid cell depend on the values of the nearest surrounding station found in 8 sectors. The stations are weighted in relation to the distance and climatological comparability (distance to sea) to the grid cell. Interpolated values will never exceed the values of the original data the grid network was calculated from. This ensures that "false" records will never appear from the interpolation. This is done for all 1x1km cells covering Denmark’s land area. The 1x1km grid is the calculation basis for all of the spatial resolutions described below. The observation network is embedded into this dataset, i.e. the interpolation algorithm is made so that the 1x1 km grid cell that geographically contains a station, always will have the observed value. The 1x1 km grid net serves as the basis of calculation for the other spatial resolutions.


Spatial Resolutions

The 1x1km grid is then used to calculate a single value for each spatial resolution (e.g. one municipality) from the grid cells covering the specified area combined with a weighting of how much of each grid cell is within the area. Below is shown an example of the 1x1 km grid covering Holbæk Municipality (Holbæk Kommune):

Figure 3. 1x1 grid net covering Holbæk Municipality (Kommune)


Aggregated 1x1km grid cells are used to compile datasets of different spatial resolution:

Spatial resolution

Description


10kmGridValue  

climate data in a 10 x 10km grid resolution covering Danish land areas. This matches the DKN_10km_ETRS89 grid from Det Danske Kvadratnet.


You can also see, how you can import grid identification layers into Qgis.

20kmGridValue

climate data in a 20 x 20km grid resolution covering Danish land areas. This grid resolution follows the same conventions as the other grids in Det Danske Kvadratnet, but this specific resolution does not exist.


You can also see, how you can import grid identification layers into Qgis.

municipalityValue

climate data for the Danish municipalities


countryValue

climate data for Denmark


For a detailed description of the interpolation, see DMI Technical Report 10-13 (the report is in Danish).

Data is made available through DMI’s open data service as soon as it has been calculated, but not yet quality controlled.  When the stationValue data has been quality controlled by our climatologists, the data is recalculated and will be updated in the service with any changes to data. In some instances, the data is based on additional parameters, which are not quality controlled. These parameters are agricultural parameters.

Climate data is not available for Greenland in different spatial resolutions, as an interpolation algorithm cannot be applied due to the complicated circumstances of the large geographical area, complex topography and low station density.


  • No labels