Rice is a staple agricultural crop and it is the main food product for the majority of the world’s population [1]. Rice prices are 2–2.5 times higher than wheat prices. In the Russian Federation (RF) the rice production is concentrated in the Southern Federal District (Krasnodar Krai, Rostov region), in the North Caucasian Federal District (The republic of Dagestan, Adygea, Kalmykia, Astrakhan region, the Chechen Republic) and in the Far East Federal District (Primorsky Krai) [16]. The rice zone of RF is adjacent to the northern border of the world’s distribution range of rice. The least heat-supplied region in Russia is located in the monsoon climate of the Far East — Primorsky Krai. The weather conditions of this area are characterized by great imbalance that causes significant inter-annual fluctuations in yield [8].

The gross harvest of rice in Primorsky Krai (PK) in 2013 amounted to 54.7 thousand tons [15]. The rice production in the region allows almost fully to meet the needs of the population in PK and the Far East in this valuable agricultural crop. Analysis of the dynamics and temporal variability of rice yields in PK is shown in Figure 1.

The studying of peculiarities of the spatial and temporal changes in climate and weather elements of PK has an important methodological implications not only for the scientific substantiation of rice crops allocation in the northern parts of the global ​​

Figure 1. Dynamics of rice yields in Primorsky Krai [13 ].

distribution area, but also for forecasting a rice yield, the adverse weather conditions, risk assessments, recommendations for usage of climatic and prognostic information.

Methods for climate influence estimation with the purpose of the production of agricultural crops, including rice are presented in works of foreign [7] and Russian researchers [4,5]. Agrometeorological conditions of rice fields in the Primorsky region were studied in detail and described in articles Chirkov Yu. I. , Pestereva N.M. [9, 10, 11, 12], Chernysheva L.S. [2, 3]. These authors proposed to consider the yield of rice in the form of two components. Currently there are about 100 different prediction methods and techniques. The most consistent assessment perspective crop yields (in this case rice), depending on the agro-climatic conditions expected following forecasting methods and techniques: methods of expert assessments, modeling; normative method; target forecasting; methods extrapolation [13].

Extrapolation methods are based on the immutability factors determining the development of the studied object, and the object are in the spread of regularities in the past — on his future. In other words, the essence of the method of extrapolation is to examine existing in the past and the present and the future. There are the following target extrapolation methods: simple extrapolation; analytical time series alignment; exponential smoothing; methods of harmonic scales; use autoregression equations and methods of «moving average». Such a method was designed and is described in detail in the articles of the famous scientist agrometeorologist — professor Pestereva N.M. [14]. It is very effective and at the same time simple enough in the performance of the method of averaging random variables (in our case — the temperature of the growing season rice), which gives an opportunity to assess the trend — the trend of further changes in those or other values.

Rice yield Р_t was considered as the sum of two components: non-random or trending Е_t due to the culture of farming (rice varieties, agricultural technology) and long-period variations in climate, and a random component ΔР_t, which is determined by the weather conditions of a particular year:

Р_{t =} Е_t ± ΔР_t,  (1)

where t – means time, years (figure 1).

1. Forecast nonrandom component Et is carried out either by means of a third degree polynomial by the formula

Е_t = at³ +bt² +ct + d, (2)

where t — number years starting in 1950 until 1988; a, b, c, d — coefficients of which are determined by a least squares. In a second embodiment E’t trend component is calculated from the empirical formula

Е‘_t = 0,2 [0,2P_t-5 + 0,45 P_t-4 + 0,78 P_t-3 + 1,28 P_t-2 + 2,29P_t-1], (3)

where P_t-i — the actual value of rice yields in previous years (I = 1, 2, 3, 4.5). When calculating the weighting coefficients in (3) to focus on yields of recent years, which to some extent makes it possible to take into account changes in the level of agricultural technology.

In PK, rice grows along the northern border of the global area of distribution. Weather and climatic conditions of the region characterized by high volatility and a high probability of extremely unfavorable and dangerous weather phenomena for rice: heat deficit, especially in the first half of the growing season (late May — early July). However, modern farming techniques, provided that farmers tailored climate information and forecasts with a big advance will allow manufacturers to significantly reduce the risks of rice production of this crop, as well as allow to develop a set of managerial decisions aimed at obtaining sustainable harvest rice in the monsoon climate of the Far East. According to a new federal program for the development of the Far East of Russia in the Primorsky Krai will be created modern agricultural cluster. It is hoped that the provision of agrometeorological in the coming years to regain its position and provide farmers a modern, high-tech and highly accurate prognostic information. This method is long-term forecast of crop yields has been adapted and successfully introduced at the forecast the other agricultural crops including aquaculture.