Aggregation of crop indicators

From Agri4castWiki
Revision as of 11:04, 4 February 2014 by Raymond (talk | contribs) (Administrative level 2, 1 and 0)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search


For each decade, simulated crop indicators of the EMU’s are spatially aggregated to the climatic grid cells for the production of crop indicator maps and to the administrative regions as a basis for yield forecasts.

EMU to grid

The aggregation from simulated yield on EMU resolution to yield in grid resolution is based on the weight of each EMU within the climatic grid cell as shown by the equation below. The weight is calculated as the cropspecific landcover area fraction of the EMU with suitable soils in relation to the total suitable area of all EMU’s within the climatic grid cell.


Y_{G}=\dfrac{\sum_{i=1}^n C_{E,i}*A_{E,i}*Y_{E,i}}{\sum_{i=1}^n C_{E,i}*A_{E,i}}


Where:
Y_{G}: simulated yield at climatic grid cell [kg ha-1]
Y_{E}: simulated yield at EMU level [kg ha-1]
A_{E}: landcover area of EMU [ha]
C_{E}: fraction of the EMU area suitable for the crop group [-]
n: number of EMU’s in climatic grid cell [-]

The simulated yield on EMU resolution Y_{E} is read from tables SIMYIELD_EMU (based on observed weather) and SIMYIELD_<MODEL>_EMU (based on HIS, OPE, EPS, MON, and SEA model). The relation between climate grid cells, EMU's and their crop specific landcover area A_{E} is read from table LINK_EMU_REGION_LANDCOVER. This table gives for each unique combination of GRID, SMU, REGION and CROP the landcover area. The suitable fraction C_{E} is read from table SMU_SUITABILITY. The results are written to table SIMYIELD_GRID (based on observed weather) or SIMYIELD_<MODEL>_GRID (based on HIS, OPE, EPS, MON, and SEA model).

The indicators that combine a crop development stage with an extreme weather event are calculated on-the-fly in the viewers (see Indicators available at 25x25 km grid resolution).

EMU to administrative regions

Administrative level 3

The aggregation of simulated crop indicators of EMU’s to the finest resolution of administrative regions (level 3) is very similar to the aggregation from EMU to grid. The weight is calculated as the cropspecific landcover area fraction of the EMU with suitable soils in relation to the total suitable area of all EMU’s within the climatic grid cell.


Y_{N3}=\dfrac{\sum_{i=1}^n C_{E,i}*A_{E,i}*Y_{E,i}}{\sum_{i=1}^n C_{E,i}*A_{E,i}}


Where:
Y_{N3}: simulated yield for the region at NUTS level 3 [kg ha-1]
Y_{E}: simulated yield at EMU level [kg ha-1]
A_{E}: landcover area of EMU [ha]
C_{E}: fraction of the EMU area suitable for the crop group [-]
n: number of EMU’s in level 3 administrative region [-]

The simulated yield on EMU resolution Y_{E} is read from tables SIMYIELD_EMU (based on observed weather) and SIMYIELD_<MODEL>_EMU (based on HIS, OPE, EPS, MON, and SEA model). The relation between administrative regions, EMU's and their crop specific landcover area A_{E} is read from table LINK_EMU_REGION_LANDCOVER. This table gives for each unique combination of GRID, SMU, REGION and CROP the landcover area. The suitable fraction C_{E} is read from table SMU_SUITABILITY. The results are written to table SIMYIELD_REGION (based on observed weather) or SIMYIELD_<MODEL>_REGION (based on HIS, OPE, EPS, MON, and SEA model). During the aggregation to administrative level 3, aggregated yields for some crops are duplicated for later differentiation on level 2, 1 and 0. For instance Winter wheat yield for administrative level 3 is duplicated as Triticale, Durum wheat and Winter barley (see Crop parameters for the CGMS).

Administrative level 2, 1 and 0

The aggregation to courser administrative regions (level 2, level 1 or level 0) is based on the area distribution between the ‘children’ of a specific region. Land use statistics for level 3, level 2 and level 1 regions are available in the table AGGREGATION_AREAS. The simulated crop indicators at level 3 are aggregated to level 2 via:


Y_{N2}=\dfrac{\sum_{i=1}^n C_{A3,i}*Y_{N3,i}}{\sum_{i=1}^n C_{A3,i}}


Where:
Y_{N2}: simulated yield for the region at administrative level 2 [kg ha-1]
Y_{N3}: simulated yield for the region at administrative level 3 [kg ha-1]
C_{A3}: crop specific cultivated area of a region at administrative level 3 [-]
n: number sub regions at administrative level 3 for a region at administrative level 2 [-]

During this process, the duplicates of (administrative level 3) are differentiated into different 'aggregation versions' because they have different aggregation areas. The results are added to table SIMYIELD_REGION (based on observed weather) or SIMYIELD_<MODEL>_REGION (based on HIS, OPE, EPS, MON, and SEA model). Simulated yields at level 1 and 0 are obtained in a similar way.