Meteorological yield analysis

from Wikipedia, the free encyclopedia

In the meteorological yield analysis , the influence of meteorological elements on the yield of agricultural crops is recorded mathematically and statistically .

The atmosphere is an important factor influencing the production of substances and the formation of yields in cultivated plants . Alfred Mäde has determined how much the harvests can depend on the weather for a year . According to his investigations, meteorologically related yield fluctuations of 40–50% of the average yield are possible in extreme years.

The plant yield is not only the result of direct meteorological influences, but is dependent to an even greater extent on the interplay between soil, plants and atmosphere. In particular, physical properties of the soil, the humus, nutrient and water balance as well as the biological activity of the soil are subject to changes based on meteorological influences, which can also determine the effectiveness of agro-technical measures. Because of this natural dependency of plant production, information about the influence of meteorological factors on the yield performance of crops and the effectiveness of yield-enhancing measures is of particular practical value. The analysis and quantification of the relationships between weather, climate and crop production is a prerequisite for the production process and the like. a. by

  • adapting the production measures to the weather,
  • the selection or breeding of site-appropriate crops and varieties and
  • the yield forecast

to be able to control and optimize better. Such measures are only possible when the demands of the cultivated plants on the atmospheric environment and the dependence of the yield formation on the course of meteorological factors in the individual sections of plant development are known and generalized in the form of mathematical models.

Mathematical-statistical methods are primarily used for the mathematical recording of meteorological environmental influences on the plant. The focus of such investigations is the yield behavior in relation to meteorological influences during the growing season and changes in behavior as a result of deviating fertilization measures. By applying multivariate methods, for example by applying factor analysis to the complex consideration of many related variables, a new way of studying meteorological influences on agricultural crops opened up in earlier years. Compared with proven univariate methods (e.g. correlation, multiple regression analysis), there are completely different approaches to the yield analysis.

Meteorological yield analyzes require homogeneous series of biological and meteorological observation data for as long as possible. If the influences of mineral fertilizer nitrogen are also included in the considerations, the data of a long-term field test can be used as a research basis. The results presented below are based on the data from the "Static Deficient Fertilization Trial" Bad Lauchstädt. They have already proven themselves several times as the basis for investigating the interaction between the nutrient effects and the annual weather pattern.

Database and methodology

Trial and production data

The plant yield is not only the result of direct meteorological influences, but also depends on the interplay between soil, plants and atmosphere. The effectiveness of agrotechnical measures also depends on meteorological influences. Potato yields in particular were subject to considerable annual fluctuations in the 1980s, which is why the author presented the results of his empirical studies on the influence of meteorological factors on the development and yield of late potatoes as well as on the nitrogen fertilizer effect in his dissertation .

Meteorological yield analyzes require a homogeneous series of agricultural and meteorological observation data over many years. If the influences of fertilization are included in such considerations, the data of a long-term field test can be used. For his investigations, Kanther used the data from the static fertilization test in Bad Lauchstädt in order to also be able to record the effects of mineral fertilizer nitrogen. These data have repeatedly proven to be a good basis for studying the interrelationships between nutrient effects and annual weather patterns.

The study material was the tuber and grain yields of late potatoes and winter wheat from 1925 to 1970 of the full fertilization (NPK) and nitrogen deficiency (PK) parcels of the test block supplied with 200 dt / ha of manure and the climate data recorded at the same location for many years.

The "Static Trial" Lauchstädt has existed since 1902 and is divided into 4 strikes. The crop rotation is: sugar beets, spring barley, potatoes, winter wheat. The experiment was laid out without repetition and with a plot size of 500 m². The fertilization has been changed a few times over the years. Since 1952 there has been an unchanged fertilization plan.

The test site is geographically on the edge of the "Querfurter Platt" on deep and humus-rich sandy loess loam in * state level I.

  • Land value number: 94
  • Soil form: Loess black earth on an average of up to 1.5 m deep loess layer with 1.7% C and 0.16% N
  • Climatic area: Central German dry area in the rain shadow of the Harz
  • Altitude: 110 m above sea level
  • Ground water level: 3 m
  • Average annual precipitation for 60 years: 480 mm
  • Average Annual mean temperature: 8.6 ° C

Further information on the test setup and implementation, multi-year average yields and the influence of the various forms of fertilization can be found at Rüther and Ansorge.

Meteorological data

The availability of meteorological data of important elements that reliably characterize the weather conditions at the test site is an important prerequisite for the analysis of existing environmental-yield relationships. At the test site in Bad Lauchstädt, the daily values ​​of the most important meteorological elements have been recorded since 1902 according to an observation program corresponding to the meteorological service. 30-year series of climatic data were considered sufficient for recording essential weather constellations. The daily values ​​(approx. 77,000 individual values) of important elements from the observation period 1949–1970, which were stored on data carriers for automatic processing and evaluated after data verification, served as the database.

Because meteorological elementary variables (e.g. daily mean temperature, daily rainfall, etc.) are unable to adequately describe the actual effect of the weather on yield generation, meteorological variables, which are to be regarded as an integral over several individual elements, were derived from the original data. According to Wang, the difference between the temperature maximum of one day and the temperature minimum of the following day characterizes the daily cooling process, so that the average daily cooling of a period can be calculated. The daily cooling can therefore be described as a complex parameter because it generally depends on the cloud conditions, the air and soil moisture and the air movement.

Using measured data and a special calculation formula, it was also possible to calculate the potential evapotranspiration (evaporation). It marks a day's water usage. If one subtracts the potential evaporation from the daily amount of precipitation, one obtains the climatic water balance , which shows quantitatively the change in moisture of the soil for one day. According to Wendling, the estimated transpiration values ​​in Central Europe agree best with the measured values ​​if the estimates are based on relationships that contain radiation elements. The calculation of the daily values ​​of the potential evaporation was therefore carried out with the Turc formula modified by Mäde and used for the investigations.

The statistical analyzes were thus based on aggregated data obtained from climatological time series of elementary, combination and complex variables for several development stages of the potato and wheat, which were related to the data of the biological variables (phase length, mean daily development rate, yield and N-dependent additional yield) related. The parameters used were the sums and mean values ​​of elementary and complex quantities and, due to the low information content of the mean values ​​of meteorological elements, also the frequencies of occurrence or exceeding of certain meteorological limit values ​​according to Mäde for the temporal course or distribution of the effects.

To record the regional weather influence on the development of late potatoes, the yield and the nitrogen fertilizer effect, data from selected production sites (period 1967–1974) and aggregated observation data from assigned full-time weather stations for observations over the northern and central inland lowlands in eastern Germany were used.

Phenological data

For the determination of the duration and speed of development between two phenological phases, for the ordering of the climatological material and the transformation of the original time scale (calendar data) into a phenologically normalized, registered starting dates of phenological phases (time of order, emergence, beginning of flowering, harvest) are of great importance Value. Because purely calendar-based considerations promise little success in recording the weather influence on crops. They hardly allow biologically understandable statements because they do not take into account the physiological demands of the cultivated plants and the different vegetation development every year. In addition, the cultivated plant goes through times during its development when it is particularly sensitive to certain influences. However, the future harvest yield is highly dependent on its course. On the basis of the annual phenological dates, typical sections of the plant development can be defined, according to which the data processing of the meteorological material can take place. The following applies to late potatoes:

  • Rise phase: order until rise
  • First growth phase : predominant herbaceous growth in root crops (emergence to beginning of flowering)
  • Second growth phase : predominant tuber formation (beginning of flowering up to the 52nd pentad)

Furthermore, the following stages of the potato can be defined:

  • T1 - 25 days from the order date (preliminary stage)
  • T2 - 25 days from emergence of the potato (emergence stage)
  • T3 - 25 days before the start of flowering (pre-flowering stage)
  • T4 - 40 days from the beginning of flowering (stage of tuber formation)
  • T5 - 40 days before the start of harvest (pre-harvest stage)

Missing observations can be replaced by phenological dates of certain wild plants. It is also possible to estimate them from weather data. For the Lauchstädt location and for the northern and central inland lowlands in eastern Germany, estimation functions were derived with which the length of certain development sections and thus the phenological start dates of the late potato can be reliably determined.

Statistical methodology

For the mathematical recording of meteorological environmental influences on a field crop, primarily statistical methods are available . The focus of the investigations carried out in the 1970s was the yield behavior in relation to meteorological influences during the growing season and the behavior changes in the event of deviating fertilization measures.

By applying factor analysis for the complex consideration of many interrelated quantities, a new way of studying meteorological influences on crops was embarked on. In studies with proven univariate methods (correlation analysis, multiple regression analysis ) and multivariate factor analysis , completely different approaches were compared.

The analysis of yield-generating factors and the environmental demands of the crops presupposes a careful selection of the influencing variables, since certain requirements (sample size, independence of influencing variables, etc.) must be met when using statistical methods. Various methods are available for determining and selecting potential variables, e.g. B.

On the one hand, it is necessary to take a complex look at the meteorological influences on yield formation, on the other hand, regression analysis is the most frequently used method. This is made more difficult by the often existing mismatch between the scope of meteorological data and the small number of yield values ​​as well as the additional problem of multicollinearity. However, with the help of factor analysis, the existing relationships between the target variable and the numerous influencing variables can be elegantly elucidated. When using factor analysis, it is assumed that close relationships between random variables or between the target variable and the influencing variables are the result of common development conditions. If one uses the measured quantities, one can deduce from them the common conditions of origin. Your task is to determine groups of characteristics (so-called factors) that are independent of a number of influencing variables and that explain how the statistical relationships came about.

Weather effects on crops in the long-term field test

Earnings development

On the basis of the average yields determined for several periods (period 1925–1970) it was found that potatoes and winter wheat in the static fertilization experiment Lauchstädt react to nitrogen deficiency fertilization (PK fertilization) with significantly lower yields. On loess black earth , with continued insufficient nitrogen fertilization, the average yield for potatoes is 25% and for wheat 15% lower than with full fertilization. The average N fertilizer effect over the years is 80 dt / ha for potatoes and 7 dt / ha for wheat. In the case of potatoes, the Lauchstadt site conditions show a clear yield-stabilizing effect of the fertilizer nitrogen. Provides information about the achieved average yields and the yield variability of the two crops in the observed periods of the period 1920–1970 with full and nitrogen deficient fertilization.

When using the long-term test data, it had to be taken into account that the annual fluctuations in yield are not only due to the different course of the annual weather, but also to changes in the test technology and the fertility of the soil. The latter influences are reflected in the earnings trend. In order to eliminate the effect of these non-meteorological influences on yield, the 45-year time series must be adjusted for the trend. The development of the harvest yields of both crops with full fertilization and nitrogen deficiency fertilization in the "static test" is shown graphically in Figures 5.1–5.4.

If the yield development is first described by a linear function, the functions shown in Table 5.2 result from the trend calculations.

They show the annual fluctuations in the tuber or grain yield with full and nitrogen deficient fertilization in the period 1925–1970 around the trend of the yield development, in which the non-meteorological influences are absorbed. The trend can be determined by adapting a 2nd degree polynomial to overlapping 7-year sections of the time series with the aid of the lubricant method (dash-dotted curve) and a compensating cubic spline function (dashed curve). Both methods lead to a matching trend adjustment. The undulating course of the trend curve cannot be overlooked. Whether this wave motion is due to existing solar-terrestrial relationships, i. H. the influence of sunspot periods with their effects on individual meteorological elements (e.g. temperature, air pressure, precipitation) and large-scale circulation should be investigated.

The diagrams describe a continuous decline in PK yields and the trend for the period 1930–1945. Only after 1945 is there a gradual increase in earnings until 1959, when the previous earnings level is reached again.

The trend is switched off by forming the yield differences between the empirical yield values ​​of the time series and the estimated trend values, which are shown as polygons (see Figure 5.9).

The absolute yield deviations mainly characterize the extent of the annual weather impact on potatoes. For meteorological yield analyzes, according to Mäde, the deviations of the annual yields are calculated from the ΔE (t) in% of the respective trend value, which can be related to the meteorological values ​​as yield values. The trend deviations indicate the different effects of the field crop against meteorological influences. During the entire study period they fluctuate in the "static test" for potatoes between -33… +38% (with NPK fertilization) or -67… +66% (with PK fertilization) and with wheat between -27… +30% ( NPK) or −32… +35% (PK). The trend deviations determined enable the formation of yield classes in good, medium and bad harvest years to differentiate the effects of the weather.

Warmth and water enjoyment of the crops in years of extreme yield

Temperature and precipitation are generally determining bioclimatic variables, because they also reflect the directly biologically effective factors such as radiation , moisture and evaporation . In order to gain an initial overview of the weather conditions during the vegetation for yield formation, years with high, medium and low harvest yields combined with temperature and precipitation can be compared in thermopluviograms (temperature sum and precipitation sum curves ).

The diagram in Figure 5.10 illustrates the weather patterns for several maximum yield years for potatoes. The temperature sums are plotted as the abscissa and the values ​​of the precipitation sums as the ordinate. After connecting the coordinates, following the course of time (potatoes 20th – 40th pentad), the curves shown result, which also implicitly contain the time magnitude. High increases in temperature are characterized by the curve stretching to the right and periods of precipitation by a steep rise. Dry seasons make horizontal curves clear. The start dates of the phenological phases, emergence and start of flowering of the potatoes are marked on the curve. Climatic diagrams enable a quantitative comparison of the heat and water consumption during the growing season of potatoes and winter wheat in good, medium and poor yield years. The following climatic diagram shows the weather patterns for the maximum yield years 1931, 1940 and 1954 for late potatoes. In contrast, the following climatic diagram in Figure 5.11 results for selected years of poor yield for potatoes (1936, 1949, 1953).

For potatoes it was found that good and bad harvest years are characterized by quite different weather patterns. Above-average crop yields were recorded in both dry and wet growing seasons. Particularly clear differences between the annual groups with regard to the development time as well as the heat and water consumption, which could have explained the strong meteorological yield differences, could not be identified with the help of climate charts.

In contrast, different temperature and precipitation conditions between the annual groups can be determined for the study period for wheat. The years with high and medium harvest yields differ from years with lower yields in that they have a significantly lower increase in temperature and precipitation totals during the pre-winter development, lower temperatures and higher rainfall in the period without vegetation and, above all, lower total increases in the vegetative development phase.

The influence of temperature and precipitation on the speed of development of wheat is very clear. While in good and medium yield years the pre-winter development lasts around 15 pentads and the dormancy period lasts 23 pentads, the opposite is true in bad harvest years. An early completion of the pre-winter development (before the 70th pentad) results in a higher harvest yield in the Lauchstadt trial.

Tables 6.7 and 5.8 summarize the extremes and means of the development period as well as the temperature and precipitation sums continuously accumulated up to the respective end of the phase for three annual groups for late potatoes and winter wheat.

With the help of a suitable water balance model (e.g. Thornthwaite, Turc, Klatt), the daily water balance in the vegetated soil can be determined for individual years using temperature and precipitation data. This provides information about the course of the daily values ​​of evaporation, soil moisture and precipitation during the vegetation and about the differences in water demand and water supply as well as water supply. On this basis, statistical observations of the water balance of certain time periods are also possible in order to determine those factors that produce differences in yield.

The application of Thornthwaite's water balance model has shown that it is possible with such a model

  • derive all significant quantities of the water balance from temperature and precipitation data so that one can relate climate-related water demand, water supply and the plant-available soil water content to plant development and plant yield,
  • to compare the water supply of different years, locations or areas;
  • To determine threshold values ​​for soil moisture, which must not be fallen below in order to avoid drought damage. For this purpose, calculations for numerous years are necessary in order to be able to conclude from the annual cycles in connection with the yield data whether the soil moisture cycle was favorable for the respective field crop or not.

However, this single-layer model (it regards the soil zone through which a stand is rooted as a coherent reservoir that contains the amount of water W (t) at time t) does not allow any statements about the course of the water content in different layers of the root zone. The model ignores the fact that a plant population draws water from different layers of soil at the same time.

Statistical analysis of weather-yield relationships

In studies, factor analysis has proven to be quite efficient, because it enabled a complex consideration of the numerous interdependent quantities in one calculation without having to forego the advantages of correlation and regression analysis . Using the principal component analysis by Hotteling, significant factors were recorded and extracted from correlation matrices as certain properties that act independently of one another. It enabled a quick clarification of the inner connections and the proof of those influencing variables and groups that determine the plant development and the plant yield as well as the fertilizer effect. The main axis transformation resulted in ranking of the influencing variables according to their direct effect on plant yield and fertilizer effect. Rotated factor structures form the basis for quantifying the determined factors.

By comparing the parameters that characterize the duration and speed of plant development and meteorological parameters, statements about the demands of the cultivated plants in individual vegetation sections can be obtained.

By means of factor analysis it could be shown that behind 18 given parameters 4 common factors, i. H. unobservable meteorological properties. These summarize the effect of several parameters on the speed of development and the crop yield. There are close relationships between the variables of plant development and the sum of the daily mean temperature, the sum of the sunshine duration, the number of days with precipitation <3 mm and sunshine> 5 hours and potential evaporation. Another group of correlated characteristics are the parameters of plant development, mean daily maximum temperature, mean daily cooling and days with precipitation <3 mm and sunshine> 1 ≤ 5 hours. Accordingly, water-sapping conditions delay the emergence of the potato plants. In the time between emergence and the beginning of flowering, the young potato plants react to warmth and drought with slower development, which has a yield-increasing effect. After flowering, sustained periods of warmth cause greater water losses, which delay plant development, which does not affect the later yield.

Estimation functions derived from regression analysis can be used to determine the speed of development of the late potato and the length of the intermediate phases (development time) from temperature and precipitation data for the Lauchstadt location.

To analyze the yield behavior on changing weather influences during the vegetation period, one can first determine the temporal course of the correlation between the yield and a meteorological parameter according to Mäde in order to find correlation extremes. The observation material of a meteorological parameter, which is to be classified according to phenological phases, is to be correlated with the deviations of the annual yield from the trend in% for overarching 5-day periods. The calculated coefficient is assigned to every third of the five days. Figures 5.25–5.29 show the correlation coefficient between the potato yield with full fertilization (NPK) and with nitrogen deficiency fertilization (PK yield) and selected meteorological parameters in the Lauchstädt permanent field test.

Figures 5.28 and 5.29 show the course of the correlations between the yield and the number of days with> 3.0 mm of precipitation or the sum of the sunshine duration based on a phenologically standardized time scale as curves. The time in days is plotted on the abscissa, starting from the time of cultivation to the time of harvest, and the correlation coefficients are plotted on the ordinate. The course over time of the coefficients calculated for individual vegetation sections can be read from the curves. The phase-related curves are assigned in such a way that the mean start date of the respective phenological phase coincides with the zero day of the phase-related correlation. The r-values ​​exceeding the 5% significance threshold are an indication of significant relationships.

The course of the correlation coefficients between the yield of winter wheat and meteorological parameters (frequency of days with daytime temperature and precipitation within specified limits, precipitation total, sunshine duration) is shown in Figures 5.30–5.33. It should be noted that this method only allows the elucidation of simple connections.

As the correlation extremes show, an increasing frequency of precipitation (> 3 mm) before the emergence of the potato plants has a favorable effect and between the 150th and 165th calendar day has an unfavorable effect on the yield. The negative correlation extremes in this period are more pronounced with nitrogen deficiency fertilization (PK) than with full fertilization (NPK). Towards the end of the vegetation period (255th – 260th day) there is an increased need for moisture. The influence of sunshine plays a particularly important role in the time between the cultivation and the start of flowering. While in the pre-emergence stage, above-average solar radiation has a negative effect on the yield. Significant differences in the temporal course of the correlations cannot be determined with full and nitrogen deficient fertilization.

In summary, it can be stated that the late potato yield in Bad Lauchstädt is primarily dependent on the influence of meteorological factors during the pre-emergence and emergence stage, whereby it reacts much more sensitively to the effects of the weather under nitrogen deficiency conditions than with a complete nutrient supply.

Simple correlations between the deviations of the annual potato yield from the trend and the meteorological parameters for three vegetation sections with full or nitrogen deficient fertilization and the proportions of the meteorological influencing variables in the multiple determinacy of the regression and the proportion of the influencing variables in the variability proportion recorded by regression can be found in Tables 5.22–5.24.

The factor-analytical investigations, including relevant variables for time periods S (1) ... S (3) of the vegetation period, led to the result that behind the 20 influencing variables there are 5 or 4 independent factors that account for 74% (full fertilization) and 84% (nitrogen Under-fertilization) explain the yield variability. While with full fertilization the weather in the pre-emergence, emergence and pre-harvest stages is important for yield-building, only the first two development stages of the potato play an essential role with nitrogen deficiency fertilization. [A time-differentiated factor-analytical consideration of the influence of meteorological parameters on the potato yields with full and insufficient fertilization in the time segments t (1)… t (5) produced the results summarized in Tables 5.34–5.38. The key figures contained there (linear correlation coefficients, rotated factor loads and communalities) allow conclusions to be drawn about the type and strength of the yield influence of meteorological parameters in the respective time period t (i). The quantities that affect income also include the order date and the duration of vegetation.

A nine-part regression function was derived from 11 selected meteorological parameters. The regression explains 95% of the meteorological variability of the deviations in potato yields from the trend. The results of linear regressions .

Weather effects on crops in the northern and central lowlands of eastern Germany

Stochastic relationships between weather and crop yield have already been worked out several times from the results of field tests. The relationships found between meteorological factors and plant development and yield under the influence of nitrogen fertilization allow insights into the interaction and effectiveness of the examined parameters, but the knowledge gained cannot be transferred to larger areas. It was therefore necessary to consider the influence of meteorological factors on crops and the fertilizer effect in larger regions on the basis of survey data at production sites and data from assigned climate stations.

Analyzes at the district or district level make little sense because these spaces are ecologically inconsistent. Due to their purely political delimitation, it must be taken into account that regionally different complexes of natural factors that influence plant growth have an effect. In numerous regions there are considerable differences in orographic relief, in the type and type of soil and in terms of climatological and hydrological conditions (Seyfert, 1962). Due to the close interaction of natural phenomena, all the essential factors that control plant development should be included in the analysis. Meteorological yield analyzes at regional level should therefore be carried out on the basis of natural landscapes. A more suitable basis is Seyfert's natural division of the area of ​​the former GDR into phenological areas.

Of the geofactor complexes, the climatological factors are to be assigned a dominating importance, because phenological differences in the individual years can mainly be traced back to their weather conditions, while other geofactors, such as surface shape, soil type and soil type, water distribution etc., remain relatively constant over longer periods of time. For this reason, the knowledge gained about the influence of weather on plant production and fertilization effectiveness relate to areas with a uniform large-scale climate, which are based on a regional division according to Böer.

This classification is based on

  • objectively determinable parameters for the climate, such as annual fluctuations, annual rate and certain threshold values ​​of the individual climate elements and
  • Experiences about the effects of the climate on landscape and vegetation.

On the basis of differentiated analyzes of the annual course of the air temperature (monthly mean) and the precipitation (sums of unequal time periods of the year) Böer has summarized groups of neighboring climate stations with the same behavior, resulting in the regional division shown in Figure 54 with different large climates. The arithmetic means calculated from the normal values ​​of the monthly mean temperature and the monthly sums of precipitation of all stations in the relevant main area characterize the characteristic differences of the major climate (Table 5.67).

The schematic climatic diagrams shown for 6 main areas are based on these values ​​(Fig. 5.41). A point in the two-dimensional parameter space corresponds to each pair of values ​​for the monthly mean temperature and the monthly total of precipitation. The points follow the course of the year and are connected by a line. According to this, the lowlands (area 1–4), low mountain range (area 5) and low mountain range (area 6) differ significantly from each other, the climatic differences between areas 2 and 3 of the inland lowlands are much smaller. The winter and summer values ​​differ only slightly from each other, so that a differentiation results from the different temperature and precipitation conditions in spring and early summer as well as in autumn and early winter.

In the main areas 2 and 3 (more maritime and continentally influenced inland lowlands) are the potato skin cultivation areas in eastern Germany, which is why these areas were particularly interesting for agro-meteorological studies.

Location selection and database

In order to obtain agro-meteorological regional information on the basis of operational records, a classification of the operational locations is essential, because this eliminates natural, agrotechnical and economic factors or complexes of factors that directly or indirectly influence the operating result and appear as systematic disturbances in the analysis. To eliminate the influence of the climate, only those locations in an area were selected that are exposed to the same large-scale climate. Table A1 in the above PDF document contains a compilation of the locations under consideration with the assigned climate and pheno stations.

When considering large climate areas, it should be noted that other geofactors also have different secondary effects in them. So z. B. Soil climates develop very differently depending on the quality of the soil under the same climatic conditions and accordingly influence plant growth. While clayey and humus-rich soils are thermally extremely inert due to their high water-binding capacity, sandy soils with their high water permeability react extremely quickly to changes in temperature, e.g. when air masses with different temperatures flow over them or due to strong irradiation or radiation. By grouping the locations according to the quality of their soil, the influence of the soil climate was partially eliminated.

Production success depends no less on the type and timing of agrotechnical measures (choice of variety, soil cultivation, fertilization, plant protection and care measures, etc.) and, last but not least, on economic factors. In order to eliminate these influences, too, the selected farms in a climatic area had to be numerically classified according to several natural, agrotechnical and economic criteria. The following were used as suitable classification features:

  • Height of the location above sea level (m)
  • average number of fields
  • N-effort (kg (ha LN mineral fertilization)
  • Stocking of cattle (GV / 100 ha LN)
  • N expenditure for potatoes (kg N / ha LN)
  • Proportion of potato area in LN (%)
  • Workforce (AK / 100 ha LN)
  • Motor horsepower stock (MPS / 100 ha LN)
  • Production fund equipment (DM / ha LN)
  • Gross production (GE dt / ha LN)
  • Potato production (dt / ha LN)
  • Mean deviation of the late potato annual yield in% from the mean
  • mean deviation of the annual gross production in% from the mean

The mean values ​​of these characteristics compiled in Tab. A45 are based on company data for the period 1967–1974.

In order to ensure the greatest possible objectivity and reproducibility, classification was not carried out in a targeted manner as usual, i.e. according to logically given classes, but numerically with simultaneous consideration of several operational characteristics.

The starting point for the classification was n locations, which should be divided into groups based on their similarity. The similarity of the locations can be determined on the basis of k features, so that the group formation is based on a matrix in which the feature values ​​are given for each location. Each location i is thus characterized by a data vector. Using a suitable similarity measure, the similarity or dissimilarity for all pairs of locations can be determined from this matrix. A principal component analysis with oblique rotation of the extracted factors was applied to the similarity matrix consisting of correlation coefficients. The most important factors each represent a group of operationally similar locations. In this way, the following groupings of locations resulted.

Once the locations had been typified, the companies were better able to be compared with regard to their production requirements. For meteorological yield analyzes, the yield data from 14 locations could then be used together with the observation data from neighboring climatic and phenostations (Tab. A1).

  • Inland lowlands with a stronger maritime influence (main area 2)

Group A:

  • Plate-Banskow; Priborn; Tuchheim-Paplitz (Mecklenburg Lake District Havelland)

Group B:

  • Satow-Kogel; Jakobshagen; Hobeck-Letzkau (Mecklenburg Lake District)
  • Inland lowlands with greater continental influence (main area 3)

Group A:

  • Jänickendorf; Worin-Seelow; Müncheberg-Heinersdorf; Golßen; Dessau-Mildensee (Central Brandenburg plates)

Group B:

  • Wurzen-Bennewitz; Wachau; Ostrau (Leipzig lowlands Lommatzscher care)

The analysis of the regional weather influence on development, yield and nitrogen fertilizer effect in late potatoes is based on the following results and effort data from farm records from the survey period 1967–1974:

  • average tuber yield (dt / ha)
  • average additional yield through nitrogen fertilization (dt / ha)
  • average production value of mineral fertilizer nitrogen (kg / kg N)
  • average nitrogen fertilizer application (kg / ha).

The key figures collected for the period 1967–1974 come from 51 selected large farms that, together with research institutions of the former Academy of Agricultural Sciences of the GDR, have carried out a large-scale production experiment since 1966 on questions of the economic use of increasing amounts of nitrogen under production conditions (Schnee et al.) . On this basis, a research contract for the quantification of weather elements and intensification measures on yield and nitrogen fertilizer effect in plant production by Kanther and Hartung (1978) was based.

It is well known that at least 20 years are required in order to reliably record the various weather constellations. The data material, which was only eight years old and was not always complete, therefore only permitted a local examination to a limited extent. In order to gain initial knowledge about the influence of meteorological parameters on development, yield and fertilizer effect under production conditions, the average annual yields or N-dependent additional yields from comparable farms within certain geographical regions were combined into yield series, which were then compared with the climate data.

In order to determine the average nitrogen fertilizer effect, the average late potato yields generated by nitrogen fertilization were used. Their calculation is based on empirically derived nitrogen fertilizer effect coefficients, with which the proportion of the total yield achieved by the amount of mineral fertilizer used is determined. Such a procedure was necessary because the so-called zero yields from the farms, ie the harvest yields achieved without nitrogen supply, were not available as an important output variable for determining the nitrogen fertilizer effect.

The very short yield series of the farms did not allow a trend adjustment of the yields. For this reason, non-meteorological impact factors were eliminated by numerical classification of the operating locations on the basis of characteristic features for the production conditions, so that they cannot appear as systematic disturbance variables in the analysis of weather-yield relationships. The statistical measures in Table 5.68 of the above PDF file characterize the late potato yield, the fertilizer effect and the nitrogen expenditure for both climatic areas and location groups.

Meteorological influences on yield and fertilizer effect

The results of the analyzes of the influence of meteorological factors on development, yield and nitrogen fertilizer effect on potatoes for areas of the inland lowlands in eastern Germany with a uniform large-scale climate on the basis of production indicators are summarized as follows:

  • In order to obtain information about the area, a classification of the location to eliminate climatic and non-meteorological factors that influence the production result appeared to be urgently required.
  • By choosing a location on the basis of areas with a uniform large-scale climate according to Böer and later
  • Group formation according to arable and plant cultivation as well as economic criteria tries to eliminate these disturbing influences.

The statistical test results presented in the tables are based on key operating figures from locations in the inland lowlands, which are more strongly influenced by maritime and continental influences.

In terms of the annual mean, the areas differ as follows:

  • Main area 2
    • mean height above sea level 44 m
    • mean air temperature 8.4 ° C
    • mean total precipitation 579 mm
  • Main area 3
    • average height above sea level 108 m
    • mean air temperature 8.5 ° C
    • mean total precipitation 586 mm

Table A1 shows a compilation of the considered regional locations with the assigned climate and pheno stations. The phenological start dates and the length of the development phases of the late potato in the inland lowlands that are more strongly influenced by maritime and continental influences (main areas 2 and 3) are shown in Table A47.

In order to obtain agro-meteorological area information on the basis of operational records, a classification of the operational locations is essential, because natural, agro-technical and economic factors or complexes of factors are taken into account. Tables A45 contains a list of key operating figures for the large plant production operations selected for the respective climatic area. On the basis of different combinations of key operating figures, the 19 operating locations were gradually classified into two groups, from which a final assignment of 14 locations emerged. The result of the location classification using the principal component analysis is shown in Table A46.

The analysis of the connections is based on eight-year data material (period 1967 ... 1974) and was carried out both on the combined material of all locations of a climatic area and separately on the material of individual location groups. Tables A48-A53 contain the extreme values, arithmetic mean and weighted mean deviations from series of meteorological and phenological characteristics, grouped according to development phases, location groups and climatic areas. The abbreviations used for the meteorological codes correspond to those listed in Table 5.13. The abbreviations D and 1 / D stand for the development time in days in the respective pheno section and 1 / D stands for the development speed of the late potato.

The results of the statistical comparison of the regional mean values ​​of the harvest yield, nitrogen-related additional yield and generation value of two location groups or climatic areas are compiled in Table A54. Likewise the results of the comparison of mean values ​​of meteorological characteristics of two location groups within the respective main area and between the two main areas.

Listed shares of the meteorological influencing variables in the multiple determinateness of the yield regressions can be found in Tables A58-A60. The proportions and determinations marked there show which of the selected parameters have a significant influence on the level of the harvest yield and the additional potato yield with nitrogen fertilization in the areas examined.

For the inland lowlands influenced by the sea (main area 2), Tables 5.69 and 5.70 summarize the simple correlations between one biological variable (phase length D, development speed 1 / D, crop yield E and additional yield from nitrogen fertilization EN) and several meteorological parameters for three phenological sections .

For the continentally influenced inland lowlands (main area 3), simple correlations can be found in Tables 5.76–5.78 for the two location groups. Factor structures resulting from factor analyzes with their charges and communalities on the influence of meteorological parameters during the vegetation period on the potato yield E and the nitrogen fertilizer effect E (N) can be found in Table 5.82.

For both climatic regions, the speed of development of the late potato in the vegetation sections under consideration (cultivation - emergence, emergence - beginning of flowering, beginning of flowering - harvest) is closely correlated with the sum of the daily mean temperature and the potential evapotranspiration.

The emergence of the potato plant is accelerated by high daily mean and daily maximum temperatures with high potential evaporation rates. In the first growth phase (emergence - beginning of flowering) the total rainfall and the dependent water balance influence the pace of development. Higher precipitation amounts during the second growth phase (beginning of flowering - harvest) inhibit plant development. In the middle inland lowlands the first growth phase is on average 5 days shorter than in the northern lowlands.

Influences of meteorological factors on the potato yield and the nitrogen fertilizer effect can only be seen in both climatic areas for the second growth phase. The meteorologically related yield spread can be traced back to the effect of 3 or 4 factors. The variability of the yield includes a moisture factor that summarizes the effects of the characteristics of precipitation, relative humidity and water balance, 22% (main area 2) or 1% (main area 2 and 3), a development factor of 14 or 6% and daytime temperature factor 11.5 or 32% stake.

From this it can be concluded that the moisture and development factor is more important in the inland lowlands, which are more strongly influenced by maritime influences, than in the continentally influenced lowlands. This finding is also made clear by the factor structures determined for the yield influence of meteorological parameters in Table A61. Factor-analytical considerations resulted in yield regressions for optimal combinations of influencing variables for both climatic areas (see Tables A62 and A63).

The parameters daily maximum temperature, potential evaporation and climatic water balance after the beginning of flowering have the greatest influence on the yield . While increased temperatures, which promote water release, reduce the yield, it increases with a supply of moisture.

The effect of nitrogen fertilizers depends on three factors. High daytime temperatures before emergence, a long initial growth phase and little rainfall after flowering result in a low effect of nitrogen fertilization. In contrast, it increases with the length of the second growth phase. During this time, water-conserving weather conditions with daytime temperatures below 16 ° C have a positive effect on the nitrogen fertilizer effect.

There is a positive, highly significant relationship (r = 0.78) between the harvest yield and the additional yield from nitrogen fertilization, which can be quantified by linear regression. With a certainty of 53%, if the production value increases by 1 kg tubers / kg N, the hectare yield increases by 200 kg, which means that in the northern inland lowlands high late potato yields require a good nitrogen fertilizer effect.

For the northern inland lowlands, seasonal weather influences on yield and fertilizer effect were also found. It should be emphasized that the nitrogen-dependent share of the yield decreases by 148 kg / ha with every above-average mm of precipitation in spring.

After working out the essential relationships between harvest yield and weather conditions for larger climatic regions of the East German inland lowlands, an attempt was made to take a local view based on the available observation material. The results obtained by factor analyzes for four locations are compiled in Tables A64-A68. They clarify the variable variance recorded by the extracted factors, the proportions of the examined characteristics and factors in the yield variability and the contributions to the yield commonality (Table A64). The other tables show the factor structures for the influence of meteorological parameters during the growing season at the individual locations.

For computer simulations, two locations each from the observation period 1968–1974 were selected from both climatic areas, the Plate-Banskow and Satow-Kogel locations from climatic area 2 and Müncheberg and Ostrau from climatic area 3.

Because the number of repetitions was too low, a biometric analysis of the influence of environmental factors that determine the yield during various development phases could not be considered. After all, it would take many years for long-term observation series to be available for statistical processing.

A fundamental prerequisite for determining the influence of growth-determining environmental factors is that the development of the plant population is observed under as different weather patterns as possible, which should fluctuate within wide limits. Experimentally, a shortening of the time required for the acquisition of observation data can be achieved by means of sowing time experiments. Mäde and Günther used this method for experiments with summer cereals. The necessary data material for the yield analysis can also be provided by computer simulation.

On the basis of the available observation material, the influence of simulated weather patterns on potato yields was examined for selected locations in the inland lowlands . The necessary data material was provided using the Monte Carlo method . It enabled the generation of normally distributed random numbers that were included in the analysis as target variables and influencing variables .

From the factor structures, the yield value of the meteorological influencing variables was determined for each operating location. The optimal combinations of influencing variables derived from these structures provided significant measures of relationship with a coefficient of determination of 66… 95%. On the basis of calculation experiments, the meteorological conditions for certain yield classes were quantified and assessed using the example of a location.

Individual evidence

  1. a b A. Mäde: Investigations on the meteorological yield analysis . In: Journal of Meteorology . No. 25 , 1975, pp. 6-16 .
  2. R. Koitzsch: In: Journal for Meteorology. No. 27 (1977), pp. 203-301.
  3. Peter Kanther: Comparative studies on the quantitative analysis of atmosphere-yield relationships, illustrated using the example of the fruit types late potato and winter wheat - a contribution to the meteorological yield analysis . In: Dissertation to obtain a doctorate in the science branch of agriculture . Martin Luther University , Halle 1980, p. 680 .
  4. Peter Kanther, Comparative Investigations on the Quantitative Analysis of Atmosphere-Yield Relationships, Illustrated Using the Example of the Fruit Types of Late Potatoes and Winter Wheat - A Contribution to Meteorological Yield Analysis , Martin Luther University, Dissertation on Obtaining a Doctoral Degree in Agriculture, Halle, 1980
  5. H. Salk: Investigations on the influence of the weather factors temperature and precipitation on the development and the yield of winter wheat in Lauchstädt. In: Dissertation, Academy of Agricultural Sciences of the GDR, Berlin. 1967.
  6. H. Rüther, H. Ansorge: Half a century "Static attempt" Lauchstädt. In: Journal for agricultural experimentation and investigation. Volume 5, (1959), pp. 99-121.
  7. ^ J. Wang: Agricultural Meteorology . Pacemaker Press, Milwaukee, Wisconsin 1963.
  8. U. Wendling: To measure and estimate the potential evaporation. In: Journal of Meteorology. No. 25 (1975), pp. 103-111.
  9. A. Mäde: For the meteorological yield analysis. 1. Sowing time trial as an investigation method. In: Contributions to social intensification and complex rationalization in plant production. 1972, pp. 197-201.
  10. A. Mäde: Agricultural climatic studies on the 100-year-old rye cultivation in Halle . Agricultural meteor. Institute University of Halle, final research report (1960), unpublished.
  11. P. Kanther: Factor analysis of atmosphere-plant relationships, shown on the fruit type potato on the basis of the results of the static experiment Lauchstädt . Lecture on the 5th Biometric Colloquium of the GDR - Region of the Biometric Society and the Biomathematics Section of the Society for Physical and Mathematical Biology of the GDR, Eisenach (1979), unpublished.
  12. CW Thornthwaite, JR Mather: Instructions and tables for computing potential evapotranspiration and the water balance. In: Publ. Climat. VOL. X, No. 3 (1957).
  13. ^ L. Turc: In: Ann. agron. 12 (1961) No. 1, pp. 13-49.
  14. ^ F. Klatt: In: Zeitschrift für Landeskultur. 8, pp. 89-98 (1967).
  15. ^ F. Seyfert In: Treatises of the Meteorological and Hydrological Service of the GDR. 8 (1962), p. 60.
  16. a b W. Böer: In: Journal for Meteorology. 17, pp. 267-275 (1965).
  17. F. Seyfert: In: Treatises of the Meteorological and Hydrological Service of the GDR. 8 (1962), p. 60.
  18. M. Schnee, Ch. Röhricht, M. Hartung, M. Reinhardt, P. Kanther: Statements about the economic benefits of nitrogen use under the conditions of specialized plant production. In: Final research report, Academy of Agricultural Sciences of the GDR, Institute for Fertilization Research Leipzig-Potsdam. 1975, unpublished
  19. P. Kanther, M. Hartung: Quantification of the influence of weather elements and intensification measures on yield and N fertilizer effect in plant production. In: Final research report, Academy of Agricultural Sciences of the GDR, Institute for Fertilization Research Leipzig-Potsdam. 1978, unpublished.