Long-term Mean Air Temperatures 1961-1990


The existence of different data bases made separate calculation of long-term mean temperatures for West Berlin and East Berlin / surrounding countryside necessary. However, the calculation procedures used were largely analogous. The procedure should be briefly outlined, since the entire procedure had to be performed for East Berlin / surrounding countryside.

The assignment of individual measurement points to continuously monitoring climate stations according to the development and other structure of their surroundings, is of central importance for various stages of calculation. It was assumed that the course of temperatures at the climate stations could be depicted at the measurement points.

As a first step, this assignment enabled the various measurements of a measurement trip to be synchronized, i.e., referenced to a point in time (von Stülpnagel 1987). The results of the measurement trips for each measurement point were then used to calculate a mean value. The long-term wind statistics for 1950 through 1970 from Berlin-Dahlem (Riemer 1971) were applied for the weighting of individual surveys: The relevant air temperature value was weighted for the determination of the mean value by means of the wind direction at the referenced point in time of each survey, and its share of the long-term mean of all wind directions.

The projection of these results onto the mean annual values for 1991-‘92 again proceeded by means of the measurement points assigned to the climate stations. The climate stations monitored the temperature continuously during 1991-‘92. This enabled the determination of mean values for any given point in time. The projection of the mean value for the measurement point required the mean annual values of the climate station and its corresponding time-referenced mean value for the measurement points. Under the assumtion that the locations of the measurement points would behave like their associated climate stations, the difference between these two values measured at the station was carried over to the associated measurement points. Figure 3 shows this procedure schematically.

Fig. 3: Schematic Diagram of Annual Mean Calculation for June 1991 - May 1992, Using two Measuring Points from a Measuring Route

Fig. 3: Schematic Diagram of Annual Mean Calculation for June 1991 - May 1992, Using two Measuring Points from a Measuring Route

The next step was the inclusion of the corresponding measurement trip results for the surrounding countryside. This first step enabled the determination of mean annual temperatures for East Berlin and the surrounding countryside for each measurement point from June 1991 through May 1992 (SenStadtUm 1993e).

Long-term Mean

A transition was now necessary to the long-term mean for 1961 through 1990, or, for West Berlin, the conversion of mean annual values for 1961-‘80 into mean annual values for 1961-‘90. This transition could be only be realized with the help of the two long-term stations at Dahlem and Tempelhof:

West Berlin

The long-term mean at both stations for 1961 through 1980 was around 0.2 °C lower than the newly calculated values for 1961 through 1990 (Institute for Meteorology, FU Berlin 1981-‘90 and DWD 1981-‘90). The adjustment of the West Berlin long-term mean to the new period proceeded on the assumption that this difference of 0.2 °C would apply for all western measurement points of the measurement network, and thus a general increase by this quantity of difference would be justified. Forest locations were the only areas where no changes were undertaken, on the basis of results provided by two stations operated in the Grunewald since 1986.

East Berlin

The calculation of long-term temperature values in East Berlin and the nearby surrounding countryside, was accomplished on the basis of the mean values calculated for 1991-‘92, differentiated according to three different use types. Climate stations were selected which presented a suitable database for the inquiry of the projection factors, because they had been in operation for several years. The Tiergarten park station was representative for parkland/ forest locations. The Alexanderplatz station was assumed to be representative for dense development. A 5-year test series between 1975 and 1980 was available for both stations. Under the assumption that these stations would fit between the Dahlem and Tempelhof long-term monitoring stations in the same manner for the period 1975-‘80 as for the long-term mean, it was possible to determine long-term mean annual values for Tiergarten and Alexanderplatz. The difference between these means and the mean annual temperatures for 1991-‘92 was then used in the projection for the associated measurement points. Figure 4 shows the principles of this procedure, using as examples the Alexanderplatz station and the measurement points for dense development.

Schematic Diagram for Calculation of Long-Term Mean, Using the Example of the Alexanderplatz Station as Reference for a Location in a Densely Built-Up Area

Fig. 4: Schematic Diagram for Calculation of Long-Term Mean, Using the Example of the Alexanderplatz Station as Reference for a Location in a Densely Built-Up Area

This process provided a long-term mean to use in projections for parkland and forest locations. It is approximately 1.3 °C below the mean annual temperature for 1991-‘92. The difference for densely developed locations was about -0.8 °C. The long-term measurement station Dahlem represented the third use type of sparse development. Its directly measured 30-year average was approximately 1.0 °C below the mean annual temperatures for 1991-‘92. Use-referenced values were increasingly included in the projection of 30-year mean temperature values for East Berlin and the surrounding countryside. The interpolation of individual values to areas was performed manually.

Update of the Data Base in 2000

The update of the map to the land use situation of 2000 was carried out in several stages. First, the results from the station and trip measurements were time-corrected. After these raw data were processed, a scaling of the temperature values was carried out. The goal was the establishment of a uniform comparative standard to permit the current data of the stations and measurement trips to be entered into the existing 1993 database. The temperature of the investigation period could be compared with the corresponding long-term values by means of the Dahlem reference station.

For areas with use structures which had changed since 1993 and could not be included in the measurement process, an update was carried on the basis of similarity analyses.