The flood forecasting is a basic work for farmland water use and regional water resources management. It plays an important role in the rational implementation of farmland irrigation drainage and the improvement of the utilization of water resources. The flood forecast is mainly a forecast of the moisture content in the field. The purpose of water conservation for the purpose of water conservation is to enable irrigation to meet the timely supply of soil water to the active root layer without generating deep leakage and wasting irrigation water. It is also necessary to minimize surface evaporation and increase soil water content. Transpiration water and transpiration water conversion efficiency.

The problem of soil water regulation is essentially to find an irrigation scheme with low investment, simple technology, and water saving. Its feature is that soil water is monitored and forecasted, and key water is poured in strict accordance with public opinion. To make effective use of irrigation water, in order to achieve the purpose of water-saving and high-yield. Therefore, in order to achieve soil water regulation, it is necessary to strengthen monitoring and forecasting of soil moisture in the field.

1 Soil moisture monitoring method and frequency in Xingtai City

When establishing local public opinion stations and regional network of stations, it must be considered that the monitoring of public opinion and drought conditions is not only a rational use of water resources, but also the most important measure and basis for achieving scientific decisions on water resources management and drought relief.

1. 1 soil moisture monitoring method (the following is the laboratory method, of course, there is now a more simple method, that is, direct measurement with the instrument, such as the soil moisture monitor or multi-point soil temperature and humidity recorder can also be used to determine .)

Soil monitoring uses the method of drying and weighing. The soil samples taken are weighed and placed in a drying oven and heated to 100 °C to 105 °C for 4 h, covered with cooling, and weighed. Formula to calculate the mass and volumetric moisture content. The soil volume, also called the soil's false weight, refers to the dry weight per unit volume of soil in the natural state of the field, usually expressed in g/cm3. With a certain volume of steel ring cutters, the soil under natural conditions is cut so that the soil just fills the ring cutter volume, and then weighed and the dry weight per unit volume, that is, the soil bulk density, is calculated according to the natural moisture content of the soil. The soil bulk density is calculated from:

Where: γ - soil bulk density (g/cm3); W soil - dry soil weight (g); V - ring knife volume (cm3).

Soil bulk density is used in addition to the calculation of total soil porosity, and can also be used to estimate soil tightness and structural conditions. Table 1 shows the test results of soil bulk density in different regions of Xingtai City.

Soil moisture can be expressed in terms of mass and volumetric moisture content. Mass moisture content: The proportion of moisture in a unit soil, without dimension, usually expressed in θm. Under natural conditions, the range of soil moisture content varies widely. For comparison, dry soil is often used as the base. Mass water content is commonly expressed as a percentage. The formula is as follows:

In the formula: θm ─ soil mass water content (%) ; W water ─ soil moisture content per unit volume (g); W soil ─ dry weight per unit volume of soil (g).

Volumetric water content: refers to the proportion of soil water per unit volume. There is no standard quantity, and it is usually represented by θv. The volumetric moisture content can represent the proportion of soil occupied by soil water, especially the volume of soil pores, which overcomes the inconvenience of the mass water content. The pore volume water content is expressed as a percentage:

In the formula: θv ─ soil volumetric water content (%); γd ─ soil dry bulk density (g/ cm3 ); other symbols are the same as above.

1. 2 soil moisture monitoring frequency

Xingtai City began to monitor droughts in 2003. The monitoring content included precipitation, groundwater depth, and soil moisture content. The soil moisture content was measured at three depths of 10 cm, 20 cm, and 50 cm, and soil water content at different depths was measured. The monitoring time is monitored once every ten days, every March-June, and September-November monitoring (starting from 2007, monitoring twice every ten days). In July and August, due to the large amount of precipitation, most of the time the soil was in saturation, and additional measurements and sampling tests were conducted according to specific conditions. Due to the climate characteristics of the north, the surface of the soil is in an icing state from December to February. The ice in the soil will reduce the infiltration of rainwater and snow melt. When the water content of ice is high enough, the soil has almost no water permeability. Frozen soil also stores more moisture in the winter, making it unable to drain or evaporate. Therefore, the moisture content in the frozen soil area during this period is relatively stable, and the soil moisture content before freezing has an effect on the entire winter.

2 Soil Moisture Prediction Methods and Models

Forecasting the growth and regression of soil moisture in the cropping layer is the basis and key to formulate a reasonable irrigation system, conduct appropriate irrigation, and increase soil water use efficiency. The process of soil moisture changes in the field is not only related to the soil properties, but also involves the exchange of moisture between the root layer and the environment (such as precipitation, irrigation, transpiration and evaporation, moisture flux at the lower boundary of the root layer, etc.). According to the existing data in the area and the reasons for observation technology, the coefficient of extinction is used. The coefficient of regression method refers to the method of applying hydrological forecasting method according to the vertical variation law of soil water to calculate the daily soil regression coefficient K value to predict the change of soil water.

Surface soil water content changes with precipitation, irrigation, runoff, infiltration, evaporation and other factors. According to the theory of rainfall and runoff and the theory of aerated water movement, for a single precipitation, the precipitation is P, the production flow is R, and the initial loss of water is I. Then the relationship between the three is:

R = PCI (4)

In the initial rainfall duration, the water balance between the surface soil and the ground is:

I = Im - Pa (5)

In the formula: Im - the largest loss in the watershed, actually represents the maximum amount of water stored in the soil, which is considered to be a fixed value in a certain area; Pa - the amount of rainfall in the early period when the second precipitation begins.

The most important factor affecting the loss (or production flow) is the amount of early-stage rainfall (ie, the previous soil moisture content). When ti = 1 day, and there is no rain on sunny days before and after, use the following formula:

Pa, t + 1 = K × Pa, t (6)

If there is rain on t, but no production, then

Pa, t + 1 = K × ( Pa, t + Pt ) (7)

In the formula: Pa ─ early influence of rainfall; Pi ─ previous rainfall; K ─ daily coefficient of soil moisture; ti ─ corresponding to Pi's rainfall from the previous rainfall; Pa, t, Pa, t + 1 ─ The time before t and the time after t affect the rainfall; Pt ─ t the amount of precipitation (mm).

When Pa, t + Pt ≥ Im, the Im value is calculated as the upper limit of Pa.

Considering the distribution characteristics of irrigation and crop roots, the target layers for the study were the sensitive layer (0-0.2 m) and the root development layer (0-0.5 m). According to the above-mentioned rainfall runoff forecasting theory, the quantitative index representing the soil moisture in the root development layer is called the public opinion index, and the formula for calculating the public opinion index can be expressed as:

Θa, t + 1 = Kt (θa, t + Pt + Qt) (8)

Where: θa, t + 1, θa, t ─ ─ (t + 1) Day and t day moisture index (mm); Pt ─ precipitation on day t (in the case of light rainfall without production); Qt - The amount of irrigation on the tth day (mm); Kt - The coefficient of soil moisture regression on the tth day.

3 Determination of coefficient K of daily regression of soil moisture content

The coefficient of soil moisture regression K is one of the important parameters in the flood forecast model. It reflects the comprehensive characteristics of soil characteristics, underlying surface, crop growth stage, water demand, and weather. According to the monitoring data of 22 monitoring stations in Xingtai City for 3 years, the regression coefficient of soil moisture content was determined. In the selected period, there is no precipitation in the two surveys and there is obvious time for retreating, select the soil moisture content θ1 at the vertical interval of the different soil depths and the soil moisture content θ2 at the end of the period, and the interval days t at the beginning of the period and the period t. Calculate the regression coefficient:

In the formula: K - soil daily regression coefficient; θ1 - average soil moisture content at different depths at the beginning of the period; θ2 - soil moisture content at the end of the period; and t - interval days. In determining the data, through analysis, data on data anomalies will not be used. This is because in actual observations, due to the use of artificial soil, there will inevitably be errors, and it may be that the area of ​​land and water causes the soil moisture content to pass. Big, it will lead to data distortion. After the analysis and screening, the purpose of removing the falsehood is achieved. Table 2 shows the analysis result of the coefficient of daily extinction of soil moisture in Xingtai City.

4 Soil moisture forecast

The forecast of moisture content is the forecast of the process of the increase and retreat of soil moisture in the root layer of crops, which is the basis for timely and appropriate irrigation. There are many factors that affect the soil moisture status, including meteorology, soil, crops, and field water management. This model forecasts from three aspects: soil moisture content, irrigation time and irrigation quota.

4. 1 Soil moisture prediction model

According to the regression coefficient of soil moisture content per ten days, the soil moisture content of (t + n) days can be predicted from the soil moisture content of t day (8). The formula is:

Θa, t + n = Knt (θa, t + Pt + Qt) (10)

Where: θa, t - calculated value of the tidal index on the tth day, mm; θa, t + n - (t + n) calculated value of the daytime somatic index, mm; Pt - predicted rainfall amount of t days, mm; Qt - The amount of irrigation water on t days, mm; the coefficient of extinction on Kt-t day, according to different time periods, take different K values.

If the forecast period spans ten days, it means that the value of K is different, and it is calculated according to different days. For example, according to the Soil Moisture Index of March 18, predict the public opinion index of March 22, and the time is in two periods of mid-March and later, the calculation formula is:

Θa, t + 5 = K31 × K22 × θa, t (11)

In the forecast period, when precipitation (or irrigation) is encountered, it is necessary to calculate the precipitation before and after the rainfall. For example, according to the soil moisture index of May 11th, the soil moisture content of May 18th is predicted; on May 15th, the precipitation is calculated by first calculating the soil moisture content of the first 5 days of precipitation on May 15, and then Soil moisture and precipitation were calculated during this period and soil moisture was calculated on May 18th.

Θa, t + 5 = K15×θa, t (12)

Θa, t + 8 = K23×(θa, t + 5 + Pt + 5) (13)

Select the monitoring data of monitoring stations in 2007, and compare the measured soil moisture data and predicted soil moisture data of different stations in different months. No precipitation was selected during this period, and the soil moisture was monitored after 5 days and the predicted soil moisture was assessed for error. The evaluation results are shown in Table 3. According to the analysis of random sampling results from 22 stations, there are 20 stations with an error of no more than ±5 %, accounting for 90.9 % of the total number of evaluations.

4. 2 irrigation time prediction model

The irrigation time model was used to predict the time when the drought occurred for irrigation. According to the soil moisture content θa, t and the minimum soil moisture content required for crop growth, θa, t + n, if the soil moisture content is lower than θa, t + n, irrigation must be performed in time, and the actual number of days during the forecast period (ie Irrigation time). The formula is:

Where: θa, t ─ day t daytime index (%); θa, t + n ─ soil moisture index (%) for maintaining crop growth; Kt 日 daily soil moisture regression coefficient; Pt ─ rainfall of day t The volume of soil moisture produced (%).

For calculation purposes, the unit of precipitation is converted to volumetric moisture in millimeters and expressed as a percentage. Calculate the thickness of the soil layer with h, then the volumetric moisture content of the rainfall calculation formula is:

Where: P ─ precipitation (mm); h ─ calculate soil thickness (mm).

If the soil thickness is 500 mm and the time period rainfall is 6 mm, the volumetric water content of this precipitation will be 1.2%.

For example: On April 16, 2007, the volumetric moisture content of Weijiazhai was measured as 22%. During the growing season, the soil moisture content was required to be no less than 21%. It is predicted that irrigation will be required several days later.

Measured soil volumetric moisture content was 22%. The daily soil retreat coefficient in mid-May was 0.989, and the time when the predicted soil volumetric moisture content fell to 21% was brought into (14):

n = (lg0. 21 - lg0. 22) / lg0. 989 = 4 (d)

To keep the volumetric moisture content of the soil not less than 21%, the farmland must be irrigated after 4 days.