CN1794292A - Remote sensing estimation method of crop multimetadata crop rotation cycle - Google Patents

Abstract

This invention relates to a remote sensing evaluating method for a multi-element rotate period of crops, which first of all selects a remote sensing image covering the crop relating to the evaluation object, then analyzes the variance on the space of the crop in the remote sensing image and takes the result as the data for evaluating the period, then sets up related formula and substitutes the analysis data to it to calculate the actual rotate period of the crop and a technology for realizing the method is related.

Description

The remote sensing estimation method of crop multielement crop rotation cycle

Technical field relates to agricultural and association area, and the estimation object is with the polynary period of crop-rotation of counting the crops of describing in big geographical regional extent.

Background technology usually will be on same plot, a kind of crop-planting is after the regular hour, plant the another kind of crop regular hour again, and the process that replaces down is called shift of crops always, this two kinds of times are called the crop rotation time from a kind of crop to another kind of crop, are also referred to as the period of crop-rotation.Crop rotation is one of main tillage and cultivation pattern of crop, has significantly reduced the use amount of agricultural chemicals, has improved ecologic environment, has higher economy, society and ecological benefits.China not only does widespread use in southern area floods and droughts and drought ring, and in northern area crop rotation also widespread use, crop rotation as corn and soybean just has tangible ecology and economic benefit, and the annual investment of country is used to subsidize the peasant household that implements soybean and corn crop rotation for nearly hundred million yuan.Worldwide, crop rotation is equally also becoming the ecology of raising agricultural production and the gordian technique of economic benefit, but improve the benefit of crop rotation, just must study the actual change of shift of crops level, and optimize the period of crop-rotation in view of the above, and how to estimate in big geographic area scope that the actual crop rotation level of crops is a difficult problem that faces in the agriculture field always, the solution of this problem is for formulating corresponding shift of crops operating strategy, have significant values, scholar both domestic and external has carried out a large amount of explorations to this.

Owing to be subjected to the restriction of technological means, for many years, the research of crop rotation periodic problem is subjected to 18, the constraint of some classical sayings in 19th century, progress seldom, it has been generally acknowledged that the crop rotation cycle is not wait in 2 to 5 years, it is bigger to dispute on, objective report is not seen in the research of the actual period of crop-rotation estimation of relevant crops as yet, the practical condition of similar saying and China is far apart on many textbooks, do not obtain the checking of science, only be the summary to experience, therefore, in fact the analytical estimating of the actual period of crop-rotation of crops is an outstanding issue by Gu so far always.

In the world, the scholar of India nineteen ninety-five via satellite image observe the crop rotation phenomenon of regional extent, but the crop rotation cycle is not studied and is estimated.The U.S. and Hesperian scholar concentrate on economy and the ecological benefits that the different crop rotation pattern of interior among a small circle test is produced to the groundwork of crop rotation research, do not estimate the actual crop rotation level of crops.

The applicant of this patent in 2004 on " Chinese cotton " magazine, publish an article first and reported the achievement that the utilization satellite remote sensing technology is estimated the period of crop-rotation of cotton paddy rice, but the evaluation method that this achievement provides has been simplified many key factors that the period of crop-rotation changes that influence, only adopted continuous 2 years satellite remote-sensing image, estimation precision is relatively poor, and its practical application has been subjected to considerable restraint.

Cause the one of the main reasons that the crop rotation cycle studies is made slow progress being: and do not know to represent how accurately and describe the crop rotation cycle, and the characteristics of crop rotation self have illustrated, the period of crop-rotation method that the single numerical value of traditional usefulness is described between the crops is unscientific, at first there is tangible symmetry in the crop rotation between the crops, secondly the crop rotation between this crops is mutual, the 3rd exists tangible asymmetry at crop rotation between the cycle, the size that is their numerical value might not equate, and all these three key features just can't be described with single numerical value at all, adopt single numerical value to be difficult to the crop rotation cycle is carried out objective evaluation, thereby restricted the crop rotation cycle is furtherd investigate, so must seek new method with scientific methods.

The objective of the invention is to adopt the remote sensing appraising new method of a kind of polynary several periods of crop-rotation, represent and describe polynary several periods of crop-rotation of crops, and by being based upon the mathematical model on statistics and the stochastic process basis, come actual polynary several periods of crop-rotation of the crops in the big geographical regional extent of analytical estimating, it is simple that it has method, science, practical value and estimation precision height such as are widely used at characteristics.

The master data of the crop rotation of different year on summary of the invention the present invention all same plot will the crop rotation zone that in big geographic range, distributes that decipher is come out from limited a plurality of satellite remote-sensing images continuous in twos on the time, according to varying in size of geographic range, it is polynary several periods of crop-rotation that unit calculates crops by city, county, township or village respectively.Therefore, the decipher result of limited a plurality of satellite images continuous in twos on the time is used for high precision and estimates that the method for actual polynary several periods of crop-rotation of crops in the big geographical regional extent becomes key character of the present invention.

The technical scheme of the remote sensing estimation method of crop multielement crop rotation cycle of the present invention is:

At first provide the key concept of polynary several periods of crop-rotation of crops, the remote sensing image that will study the crops in crop rotation zone to the covering of obtaining carries out decipher then, obtains the crop rotation data in each township or town (city or county); Use the polynary several period of crop-rotation estimation equations of crops that these crop rotation data are analyzed again, can obtain polynary several periods of crop-rotation of crops.

Crops remote sensing image of the present invention decipher mainly comprises four steps, at first raw video is carried out geometry correction, pre-service such as linear stretch enhancing; Secondly determine the interpret tag of crop rotation crops and other crops according to open-air on-site inspection; And then by the artificial visual decipher, from image, extract the information of crop rotation crops and generate the decipher map that contains crop rotation crop map spot, afterwards, this map is verified inspection on the spot, problematic figure spot to be revised, the decipher that makes map is accurately and reliably; Generate the vector quantization decipher map that comprises the crop rotation crops at last, and by Geographic Information System the decipher map is carried out spatial analysis and handle, obtain crop rotation farming rotation of crops data.

Polynary several periods of crop-rotation of research crops, usually must observe for many years the crop rotation zone, and the data that obtained are carried out statistical study, could obtain compellent result, therefore, the data that obtain from limited a plurality of remote sensing satellite images continuous in twos on the time of utilization of the present invention are estimated polynary several periods of crop-rotation of crops.Limitedly a plurality ofly be meant that at remote sensing satellite image continuous in twos on the time image of remote sensing satellite for many years per two width of cloth images on the time of choosing are continuous, but on the time, can be interrupted between continuous per two width of cloth images.

The polynary several period of crop-rotation estimation equations of crops of the present invention's design, except utilizing limited a plurality of remote sensing satellite imaging monitor continuous in twos on the time, adopted the method that the crop rotation level in all small towns in the crop rotation zone is observed simultaneously, the crop rotation level of estimating a Utopian typical small towns is (to the city, the county can be same discussion), be equivalent to expand to the random test sample size of township's one-level a plurality of, so that extrapolate polynary several periods of crop-rotation of these regional crops more exactly, it is according to being: in crop rotation and tillage and cultivation management, have tangible mutual independence between each small towns and each peasant household, on the basic coincidence statistics to the requirement of sample independence; When the periodicity of crop rotation is reflected in the shift of crops process of describing with stochastic process, this process has stationarity, the quantity that just is equivalent to annual crop rotation statistically should be roughly the same, can go out to finish needed time of whole crop rotation or cycle according to the quantity survey (surveying) of annual crop rotation; The limited a plurality of statistical nature or cycle of crop rotation process that the independent sample in continuous in twos a plurality of small towns quantitatively has township's one-level of stationarity for research on the time should be enough big.

Now, as example the polynary several period of crop-rotation estimation equations of crops being described with cotton and paddy rice wheel, is key character of the present invention but this formula is applicable to the estimation of all polynary several periods of crop-rotation of crops.

In adjacent 2 years of limited a plurality of remote sensing satellite imaging monitors continuous in twos on the time, the plot that grew cotton last one year with " cotton/rice " expression in the next year kind area of paddy rice.For given plot, if " cotton _i/ rice _I+1" be illustrated on this plot, i grows cotton and the area of i+1 kind paddy rice, then the next year kind paddy rice i.e. " area of paddy rice of ratio that accounts for 1 year area that grows cotton _I+1The area of/cotton _i" be defined as this plot i to the cotton rice crop rotation factor of i+1 or abbreviate the cotton rice crop rotation factor of this plot i as, and use CRRF _i(Cotton-Rice Rotation Factor) represent,

${\mathrm{CRRF}}_i=\frac{{\mathrm{NRA}}_i}{\mathrm{CC}{A}_i};$

CCA wherein _iBe the area that grows cotton at i for given plot; NRA _iBe CCA _iIn, the area of i+1 kind paddy rice.

Suppose only to have cotton and paddy rice in a certain given plot enterprising road wheel do, to this plot continuous monitoring N, this crop rotation process can be described with having obviously periodic stationary stochastic process, and at this point on the piece, the cotton rice crop rotation factor is CRRF _i(i=1 ..., N-1), then the period of crop-rotation CRRP of cotton and paddy rice (Cotton-Rice Rotation Period) is

$\mathrm{CRRP}=\underset{N\→\∞}{\lim }\frac{1}{N-1}\underset{i=1}{\overset{N-1}{\mathrm{\Σ}}}\frac{1}{\mathrm{CRR}{F}_i};$

Because the year number N that in fact observes is limited, so when N is enough big, the approximate value of cotton rice period of crop-rotation CRRP:

$\stackrel{\‾}{\mathrm{CRRP}}=\frac{1}{N-1}\underset{i=1}{\overset{N-1}{\mathrm{\Σ}}}\frac{1}{{\mathrm{CRRF}}_i}.$

Because cotton and paddy rice crop rotation process can be described with having obviously periodic stationary stochastic process, therefore the quantity of annual cotton crop rotation statistically should be roughly the same, and cotton is that existence is also computable with the paddy rice period of crop-rotation.

Because only have cotton and paddy rice in given plot enterprising road wheel do, so according to CRRF _iDefinition, at given " last one year of cotton area " CCA _iSituation under, if " cotton/rice " area NRA is arranged every year _iReplant paddy rice, then

Year can replant " last one year of cotton area ", so CRRP _i=1/CRRF _iSet up.

Suppose that desired cycle CRRP is the mathematical expectation E of stochastic variable ξ (ξ), the method for so approximate definite CRRP is that ξ is carried out duplicate sampling N time, produces the sequence ξ of separate ξ value ₁..., ξ _N(ξ ₁=1/CRRF ₁...., ξ _N=1/CRRF _N), its arithmetic mean then

$\stackrel{\‾}{{\mathrm{\ξ}}_N}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\ξ}}_i;$

According to powerful several theorems,

$P(\underset{N\→\∞}{\lim }\stackrel{\‾}{{\mathrm{\ξ}}_N}=\mathrm{CRRP})=1;$

Therefore, when N is fully big,

$\stackrel{\‾}{{\mathrm{\ξ}}_N}\≈E\left(\mathrm{\ξ}\right)=\mathrm{CRRP};$

The probability of setting up equals 1, and is therefore available

As the estimated value of the amount of asking CRRP.

According to above-mentioned theorem, can utilize M small towns N continuous year satellite monitoring data to the crop rotation zone, calculate the cotton rice period of crop-rotation in this crop rotation zone,

$\stackrel{\‾}{\mathrm{CRRP}}=\frac{1}{N-1}\underset{i=1}{\overset{N-1}{\mathrm{\Σ}}}\frac{1}{{\mathrm{CRRF}}_i}=\frac{1}{N-1}\underset{i=1}{\overset{N-1}{\mathrm{\Σ}}}\left(\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\frac{{{\mathrm{CCA}}_{\mathrm{ij}}}^2}{{\mathrm{NRA}}_{\mathrm{ij}}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{CCA}}_{\mathrm{ij}}}\right);$

Wherein M is the small towns sum; NRA _IjBe that j small towns is in the area that i grows cotton, at the area of i+1 kind paddy rice; CCA _IjBe j the area that the small towns grows cotton at i.

In like manner, for the situation of paddy rice and cotton crop rotation, can release the cotton period of crop-rotation formula of the rice of calculating the crop rotation zone

$\stackrel{\‾}{\mathrm{RCRP}}=\frac{1}{N-1}\underset{i=2}{\overset{N}{\mathrm{\Σ}}}\frac{1}{{\mathrm{RCRF}}_i}=\frac{1}{N-1}\underset{i=2}{\overset{N}{\mathrm{\Σ}}}\left(\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\frac{{{\mathrm{CCA}}_{\mathrm{ij}}}^2}{{\mathrm{NCA}}_{\mathrm{ij}}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{CCA}}_{\mathrm{ij}}}\right);$

Wherein M is the small towns sum; RCRF _iWhat are arranged in i-1 kind paddy rice in the area that grows cotton for i; NCA _IjBe j small towns in the area of i-1 kind paddy rice, the area that grows cotton at i; CCA _IjBe j the area that the small towns grows cotton at i.

Be generalized to generalized case, claim that the crop rotation cycle of trying to achieve with said method is the statistics period of crop-rotation of crops, abbreviate the statistics period of crop-rotation of crops as.Therefore, in big geographic range, the crop rotation of crop x and y can be described with the statistics period of crop-rotation between them, according to above-mentioned analysis, in the ordinary course of things, crop x is different with the period of crop-rotation of crop y and the period of crop-rotation of crop y and crop x, and this physical meaning in the crops statistics asymmetry between the period of crop-rotation and the common asymmetry of crop rotation between the cycle is identical.Therefore, the period of crop-rotation between crops x and the y can describe with the polynary number in following m rank:

(x ₁，...，x _m，y ₁，...，y _m，f ₁(x ₁，...，x _m，y ₁，...，y _m)，...，f _p(x ₁，...，x _m，y ₁，...，y _m))

Wherein: x _i(i=1 ..., m) be the crops x that tried to achieve with m kind diverse ways and the period of crop-rotation of crops y; y _i(i=1 ..., m) be the crops y that tried to achieve with m kind diverse ways and the period of crop-rotation of crops x; f _j(x ₁..., x _m, y ₁..., y _m) (j=1 ..., p) be reflected in period of crop-rotation x for what ask with p kind diverse ways _iAnd y _i(i=1 ..., m) between difference function, therefore the unit in polynary several periods of crop-rotation can be divided three classes, the first kind is relevant with the crop rotation of crops x and y, second class is relevant with the crop rotation of crops y and x, the 3rd class is relevant with the second class unit with the first kind.

Supposition now only is respectively x with the period of crop-rotation that the cotton period of crop-rotation of above-mentioned rice and cotton rice period of crop-rotation formula are tried to achieve between crops x and the y ₁And y ₁, x is described ₁And y ₁Between the cycle differentiation function be f ₁(x ₁, y ₁)=y ₁-x ₁, polynary several periods of crop-rotation of single order of then describing crops x and y crop rotation are (x ₁, y ₁, y ₁-x ₁).By comparing with the period of crop-rotation between the y with crops x that the single numerical value of traditional usefulness is described, polynary several periods of crop-rotation of crops provide more comprehensively information, be a kind of describing method of more science, and can obtain many important properties, with the polynary several period of crop-rotation (x of single order ₁, y ₁, y ₁-x ₁) be example: if y ₁-x ₁＞0, illustrate that then crop rotation makes the income of crops x bigger than the income of crop y, this is because if for the crop rotation of crops x and y, crop rotation is to make crop x have higher ecology and economic benefit, crops x and the length of the y period of crop-rotation have confidential relation with ecological and economic benefit usually so, and the shorter periods of crop-rotation and crops y and x of the period of crop-rotation of crops x and y are long, more favourable to crops x usually, promptly for the comparisons of two polynary several periods of crop-rotation, at x ₁Under the identical situation, more to consider difference function value y ₁-x ₁Size, promptly to y ₁Value compares, y ₁The effect of the big more then crop rotation of value is good more.

In sum, illustrate that as example the polynary several period of crop-rotation estimation equations of single order of crops are as follows with cotton and paddy rice wheel, but it is key character of the present invention that this formula is applicable to the estimation of all polynary several periods of crop-rotation of crops single order, can make similar analysis for polynary several periods of crop-rotation of the m rank of crops.Polynary several periods of crop-rotation of the single order of crops x and y crop rotation:

(x ₁，y ₁，y ₁-x ₁)

Wherein: $x_1=\stackrel{\‾}{\mathrm{CRRP}};y_1=\stackrel{\‾}{\mathrm{RCRP}};y_1-x_1=\stackrel{\‾}{\mathrm{CRRP}}-\stackrel{\‾}{\mathrm{RCRP}}.$

In the following analysis, if do not specified, all polynary several periods of crop-rotation all are meant polynary several periods of crop-rotation of single order.

Embodiment

Example

Northern Suzhou, Jiangsu Province city is the big city of famous agricultural, the whole nation, the farming of going to river in being positioned at is distinguished, 2393 square kilometres of the total areas, population 1,550,000 is had jurisdiction over 45 small towns, the existing arable land 130,000 hectares, water surface area is 1/4th of whole city's total area, be national commodity food and high quality cotton base, produce 1100000 tons in grain, 40,000 tons in cotton per year, corn, soybean, sweet potato, vegetables etc. are arranged with paddy rice, the cotton growing crop same period.Survey region is mainly cotton and paddy rice crop rotation, in addition cotton also with other shift of crops, but be not main flow wheel operation mode.

The satellite image that research is adopted is that 11937 No. 7 satellite images in land and orbit number are No. 5 satellite images in land of 11937, has covered above-mentioned survey region, and each pixel of image or grid are of a size of 30 meters * 30 meters.In survey region, cotton and cotton rice crop rotation is observed when best is that July is to August mutually to rice.Therefore, the date of the four-period TM remote sensing image that we select is July 26 calendar year 2001, on July 29th, 2002, on July 24th, 2003, on July 26th, 2004, and the quality of image meets the decipher requirement.

According to above-mentioned cotton rice period of crop-rotation formula $\stackrel{\‾}{\mathrm{CRRP}}=\frac{1}{N-1}\underset{i=1}{\overset{N-1}{\mathrm{\Σ}}}\frac{1}{{\mathrm{CRRF}}_i}=\frac{1}{N-1}\underset{i=1}{\overset{N-1}{\mathrm{\Σ}}}\left(\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\frac{{{\mathrm{CCA}}_{\mathrm{ij}}}^2}{{\mathrm{NRA}}_{\mathrm{ij}}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{CCA}}_{\mathrm{ij}}}\right),$ The data that the remote sensing image decipher obtains are analyzed, and the result is as shown in table 1.

Table 1 northern Suzhou city calendar year 2001 to the 2004 year horizontal remote sensing investigation of cotton rice crop rotation (square measure: hectare, cycle unit: year)

The small towns name	01 cotton	01 cotton/02 rice	Cycle 1	02 cotton	02 cotton/03 rice	Cycle 2	03 cotton	03 cotton/04 rice	Cycle 3	04 cotton	Average period
												The An Feng border town big buttress of prosperous flourish Chen Bao is sought greatly Da Zou Dai Nandai kiln and is swung Zhu Di buttress fishing Dong Baodong Tan Duo field Gu Zhuan	76.53 363.25 105.89 691.66 124.03 367.09 276.22 150.56 293.42 465.88 337.14 287.15 19.66 428.51 6.90 316.40	39.12 185.13 37.92 423.37 71.55 186.9 168.49 51.10 119.88 143.97 187.20 120.02 9.40 196.89 2.87 160.08	1.96 1.96 2.79 1.63 1.73 1.96 1.64 2.95 2.45 3.24 1.80 2.39 2.09 2.18 2.40 1.98	93.06 185.45 129.91 435.91 59.74 287.99 13.12 107.04 291.94 356.51 122.43 278.13 32.63 383.76 2.91 156.97	27.12 49.86 50.67 188.49 21.45 78.6 0.71 28.2 67.46 64.44 48.51 36.90 7.31 155.58 0.20 45.37	3.43 3.72 2.56 2.31 2.79 3.66 18.48 3.80 4.33 5.53 2.52 7.54 4.46 2.47 14.55 3.46	221.16 50.57 95.29 303.14 87.09 172.86 298.81 29.72 158.43 421.65 185.99 624.02 111.54 445.85 4.45 45.13	55.54 17.21 11.83 164.16 14.62 48.57 151.72 8.37 35.27 95.35 46.93 238.94 17.70 227.40 1.17 9.70	3.98 2.94 8.05 1.85 5.96 3.56 1.97 3.55 4.49 4.42 3.96 2.61 6.30 1.96 3.80 4.65	171.17 74.45 90.37 336.37 106.85 217.83 164.57 26.28 392.67 463.33 318.79 430.72 34.88 355.48 2.20 177.39	3.12 2.87 4.47 1.93 3.49 3.06 7.36 3.43 3.76 4.40 2.76 4.18 4.29 2.20 6.92 3.36

The Liu Lu Maoshan Mountain, Lin Tan Lincheng, the old country fair Li Jian of Hainan, Haihe River He Ta Red Star woods lake is given up under the Bao of west, the opinion Tang Liu Tao village, old Shen country fair and is newly piled up neatly Xu Yang Yongfeng and open fort Zhong Xu Zhouzhuang Zhu Hong in the Guo Zhao sun	233.56 172.93 248.87 853.46 691.01 622.23 74.68 247.25 637.69 701.98 317.49 228.67 471.15 383.84 490.43 340.89 96.09 515.44 128.14 77.00 473.74 497.24 112.71 303.27 453.51 183.06	131.92 63.91 117.65 479.02 418.52 260.45 38.68 82.30 430.40 439.08 178.96 87.17 286.40 185.80 179.79 147.88 44.23 312.75 43.94 25.26 255.65 259.20 52.89 200.20 301.15 100.77	1.77 2.71 2.12 1.78 1.65 2.39 1.93 3.00 1.48 1.60 1.77 2.62 1.65 2.07 2.73 2.31 2.17 1.65 2.92 3.05 1.85 1.92 2.13 1.51 1.51 1.82	119.68 163.31 268.35 816.74 726.02 257.12 50.39 255.91 554.55 647.28 241.59 248.73 426.75 334.77 408.59 241.02 66.63 421.50 202.76 138.42 346.36 315.31 11.28 373.74 163.99 130.67	18.27 18.93 67.25 424.61 279.74 26.12 18.11 55.62 278.37 307.81 98.31 55.90 203.31 120.16 119.21 39.55 16.82 156.49 50.12 30.74 171.26 140.31 4.95 150.77 82.40 26.32	6.55 8.63 3.99 1.92 2.60 9.84 2.78 4.60 1.99 2.10 2.46 4.45 2.10 2.79 3.43 6.09 3.96 2.69 4.05 4.50 2.02 2.25 2.28 2.48 1.99 4.96	364.33 346.97 208.27 853.44 943.96 136.67 93.86 205.11 715.85 611.38 136.17 193.15 422.97 227.40 273.19 103.62 247.71 486.90 135.50 100.47 302.96 197.98 156.47 564.40 100.30 196.07	174.22 66.42 50.17 467.38 382.49 65.82 17.29 30.22 281.37 299.06 70.35 40.02 174.54 96.79 51.15 43.53 79.10 225.13 21.25 13.84 154.61 122.95 67.34 238.88 53.10 54.47	2.09 5.22 4.15 1.83 2.47 2.08 5.43 6.79 2.54 2.04 1.94 4.83 2.42 2.35 5.34 2.38 3.13 2.16 6.38 7.26 1.96 1.61 2.32 2.36 1.89 3.60	208.77 214.72 310.45 638.79 1134.92 293.34 100.26 240.10 471.39 591.81 258.08 307.86 572.82 453.36 470.52 154.77 149.29 611.38 215.22 165.00 353.00 261.37 101.86 557.59 91.23 222.19	3.47 5.52 3.42 1.84 2.24 4.77 3.38 4.80 2.01 1.92 2.06 3.97 2.06 2.40 3.83 3.59 3.09 2.17 4.45 4.94 1.95 1.93 2.24 2.12 1.79 3.46
												Add up to	13866.62	7227.86	2.01	10868.96	3832.32	3.42	11580.80	4485.97	3.01	12513.44	2.81

According to table 1, it is bigger that calendar year 2001 to 2004 year is planted cotton area fluctuation, and for example, from calendar year 2001 to 2002 year, area descends 21.62%; From 2002 to 2003, area rose 6.55%; And from 2003 to 2004, area continued to rise 8.05%.The same period, each small towns to plant the fluctuation of cotton area also very big, the small towns that area increases, calendar year 2001 to 2002 year is 10; 2002 to 2003 is 20; 2003 to 2004 is 25.The number that cotton area increases is planted in the small towns, has reflected that basically Xinghua City plants the general trend of cotton area increase and decrease.From the characteristics of cycle variation pattern, according to cotton rice period of crop-rotation of the data computation of calendar year 2001 to 2002 year be 2.01 years; The data computation cotton rice period of crop-rotation according to 2002 to 2003 is 3.42; The cotton rice period of crop-rotation according to the data computation in 2003 to 2004 is 3.01; Differ greatly, but reflected basically since the calendar year 2001, particularly calendar year 2001, the influence to cotton rice crop rotation is adjusted in each small towns plant husbandry of Xinghua City.In general, by the average period that the arithmetic mean of three cycle datas of cotton rice period of crop-rotation monitoring was obtained in continuous 4 years be 2.81, only differ 12.40% with the saying of the scientist Xu Guangqi of China Ming Dynasty 2 to 3 years (on average being about 2.5 years), more approaching.

According to the cotton period of crop-rotation formula of above-mentioned rice $\stackrel{\‾}{\mathrm{RCRP}}=\frac{1}{N-1}\underset{i=2}{\overset{N}{\mathrm{\Σ}}}\frac{1}{{\mathrm{RCRF}}_i}=\frac{1}{N-1}\underset{i=2}{\overset{N}{\mathrm{\Σ}}}\left(\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\frac{{{\mathrm{CCA}}_{\mathrm{ij}}}^2}{{\mathrm{NCA}}_{\mathrm{ij}}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{CCA}}_{\mathrm{ij}}}\right),$ The data that the remote sensing image decipher obtains are analyzed, and the result is as shown in table 2.

The horizontal remote sensing investigation table of the table 2 northern Suzhou city cotton crop rotation of calendar year 2001 to 2004 year rice (square measure: hectare, cycle unit: year)

The small towns name	01 cotton	01 rice/02 cotton	Cycle 1	02 cotton	02 rice/03 cotton	Cycle 2	03 cotton	03 rice/04 cotton	Cycle 3	04 cotton	Average period
												The An Feng border town big buttress of prosperous flourish Chen Bao is sought greatly Da Zou Dai Nandai kiln and is swung the Zhu Di buttress fishing Liu Lu Maoshan Mountain, Lin Tan Lincheng, Dong Baodong Tan Duo field Hainan, Gu Zhuan Haihe River He Ta Red Star old country fair Li Jian woods lake and give up under the Bao of west, the opinion Tang Liu Tao village, old Shen country fair and newly pile up neatly Xu Yang Yongfeng	76.53 363.25 105.89 691.66 124.03 367.09 276.22 150.56 293.42 465.88 337.14 287.15 19.66 428.51 6.90 316.40 233.56 172.93 248.87 853.46 691.01 622.23 74.68 247.25 637.69 701.98 317.49 228.67 471.15 383.84 490.43 340.89 96.09 515.44 128.14 77.00	42.58 46.48 54.4 176.25 21.46 88.57 0.96 31.67 79.25 101.83 45.03 59.76 12.61 212.42 1.43 39.39 38.97 24.21 107.31 487.3 305.33 89.72 16.03 87.49 352.46 343.81 95.94 85.85 185.91 119.77 125.87 81.44 16.17 180.51 73.89 46.88	2.19 3.99 2.39 2.47 2.78 3.25 13.67 3.38 3.68 3.50 2.72 4.65 2.59 1.81 2.03 3.99 3.07 6.75 2.50 1.68 2.38 2.87 3.14 2.93 1.57 1.88 2.52 2.90 2.30 2.80 3.25 2.96 4.12 2.34 2.74 2.95	93.06 185.45 129.91 435.91 59.74 287.99 13.12 107.04 291.94 356.51 122.43 278.13 32.63 383.76 2.91 156.97 119.68 163.31 268.35 816.74 726.02 257.12 50.39 255.91 554.55 647.28 241.59 248.73 426.75 334.77 408.59 241.02 66.63 421.50 202.76 138.42	63.7 15.89 19.44 159.28 25.57 45.9 103.42 7.99 43.47 115.77 65.31 181.56 28.53 209.72 2.40 14.59 111.69 90.8 63.06 437.6 428.52 40.55 34.92 41.73 399.27 332.73 71.06 50.71 219.8 112.25 58.38 31.41 74.55 226.85 35.53 25.81	3.47 3.18 4.90 1.90 3.41 3.77 2.89 3.72 3.64 3.64 2.85 3.44 3.91 2.13 1.85 3.09 3.26 3.82 3.30 1.95 2.20 3.37 2.69 4.92 1.79 1.84 1.92 3.81 1.92 2.03 4.68 3.30 3.32 2.15 3.81 3.89	221.16 50.57 95.29 303.14 87.09 172.86 298.81 29.72 158.43 421.65 185.99 624.02 111.54 445.85 4.45 45.13 364.33 346.97 208.27 853.44 943.96 136.67 93.86 205.11 715.85 611.38 136.17 193.15 422.97 227.40 273.19 103.62 247.71 486.90 135.50 100.47	47.37 22.76 30.55 191.22 38.89 62.6 41.28 7.79 71.01 84.97 164.68 95.23 5.47 135.69 0.9 63.83 40.37 37.94 67.19 348.37 415.84 33.02 43.45 40.64 257.39 305.48 134.14 58.83 348.9 194.6 154.33 15.14 32.48 249.74 44.47 33.46	3.61 3.27 2.96 1.76 2.75 3.48 3.99 3.37 5.53 5.45 1.94 4.52 6.38 2.62 2.44 2.78 5.17 5.66 4.62 1.83 2.73 8.88 2.31 5.91 1.83 1.94 1.92 5.23 1.64 2.33 3.05 10.22 4.60 2.45 4.84 4.93	171.17 74.45 90.37 336.37 106.85 217.83 164.57 26.28 392.67 463.33 318.79 430.72 34.88 355.48 2.20 177.39 208.77 214.72 310.45 638.79 1134.92 293.34 100.26 240.10 471.39 591.81 258.08 307.86 572.82 453.36 470.52 154.77 149.29 611.38 215.22 165.00	3.09 3.48 3.42 2.05 2.98 3.50 6.85 3.49 4.29 4.20 2.50 4.20 4.29 2.18 2.11 3.29 3.83 5.41 3.47 1.82 2.44 5.04 2.71 4.58 1.73 1.89 2.12 3.98 1.95 2.38 3.66 5.49 4.01 2.31 3.80 3.93

Open fort Zhong Xu Zhouzhuang Zhu Hong in the Guo Zhao sun	473.74 497.24 112.71 303.27 453.51 183.06	219.15 158.55 7.75 195.79 78.59 56.86	1.58 1.99 1.46 1.91 2.09 2.30	346.36 315.31 11.28 373.74 163.99 130.67	164.57 99.34 48.82 257.62 61.34 80.76	1.84 1.99 3.21 2.19 1.64 2.43	302.96 197.98 156.47 564.40 100.30 196.07	156.23 123.16 22.56 236.23 59.53 62.42	2.26 2.12 4.52 2.36 1.53 3.56	353.00 261.37 101.86 557.59 91.23 222.19	1.89 2.03 3.06 2.15 1.75 2.76
												Add up to	13866.62	4595.64	2.61	10868.96	4702.21	2.70	11580.80	4580.15	3.35	12513.44	2.89

According to table 2, the cotton crop rotation of rice average period is 2.89, very approaching 2.81 average periods with cotton rice crop rotation, both only differ 2.85%, the crop rotation area of explanation from the cotton to the paddy rice is basic identical with the crop rotation area from the paddy rice to the cotton, and this phenomenon meets the basic law of crop rotation, and also meeting above-mentioned is to have periodically and the hypothesis of the stochastic process of stationarity about the crop rotation process, and illustrate that crop rotation estimation equation of the present invention is reasonably, the precision of remote Sensing Interpretation has reached the requirement of test.

In sum, polynary several periods of crop-rotation of the single order of cotton and paddy rice are calculated as follows:

Because: (x ₁, y ₁, y ₁-x ₁)

Wherein: $x_1=\stackrel{\‾}{\mathrm{CRRP}};y_1=\stackrel{\‾}{\mathrm{RCRP}};y_1-x_1=\stackrel{\‾}{\mathrm{CRRP}}-\stackrel{\‾}{\mathrm{RCRP}}.$

So: (x ₁, y ₁, y ₁-x ₁)=(2.81,2.85,2.85-2.81)=(2.81,2.85,0.04)

The polynary several periods of crop-rotation that are northern Suzhou city cotton and paddy rice are: (2.81,2.85,0.04).

On the basis of table 1 and table 2, the polynary several periods of crop-rotation that can calculate each small towns, northern Suzhou city are as shown in table 3.

The remote sensing appraising of cotton polynary several periods of crop-rotation of rice of table 3 northern Suzhou city

The small towns name	x 1	y 1	y 1-x 1	Polynary one number time
					The An Feng border town big buttress of prosperous flourish Chen Bao is sought greatly Da Zou Dai Nandai kiln and is swung Zhu Di buttress fishing Dong Baodong Tan Duo field	3.12 2.87 4.47 1.93 3.49 3.06 7.36 3.43 3.76 4.40 2.76 4.18 4.29 2.20 6.92	3.09 3.48 3.42 2.05 2.98 3.50 6.85 3.49 4.29 4.20 2.50 4.20 4.29 2.18 2.11	-0.03 0.61 -1.05 0.12 -0.51 0.44 -0.51 0.06 0.53 -0.20 -0.26 0.02 0.00 -0.02 -4.81	(3.12，3.09，-0.03) (2.87，3.48，0.61) (4.47，3.42，-1.05) (1.93，2.05，0.12) (3.49，2.98，-0.51) (3.06，3.50，0.44) (7.36，6.85，-0.51) (3.43，3.49，0.06) (3.76，4.29，0.53) (4.40，4.20，-0.20) (2.76，2.50，-0.26) (4.18，4.20，0.02) (4.29，4.29，0.00) (2.20，2.18，-0.02) (6.92，2.11，-4.81)

The Liu Lu Maoshan Mountain, Lin Tan Lincheng, the old country fair Li Jian of Hainan, Gu Zhuan Haihe River He Ta Red Star woods lake is given up under the Bao of west, the opinion Tang Liu Tao village, old Shen country fair and is newly piled up neatly Xu Yang Yongfeng and open fort Zhong Xu Zhouzhuang Zhu Hong in the Guo Zhao sun	3.36 3.47 5.52 3.42 1.84 2.24 4.77 3.38 4.80 2.01 1.92 2.06 3.97 2.06 2.40 3.83 3.59 3.09 2.17 4.45 4.94 1.95 1.93 2.24 2.12 1.79 3.46	3.29 3.83 5.41 3.47 1.82 2.44 5.04 2.71 4.58 1.73 1.89 2.12 3.98 1.95 2.38 3.66 5.49 4.01 2.31 3.80 3.93 1.89 2.03 3.06 2.15 1.75 2.76	-0.07 0.36 -0.11 0.05 -0.02 0.20 0.27 -0.67 -0.22 -0.28 -0.03 0.06 0.01 -0.11 -0.02 -0.17 1.90 0.92 0.14 -0.65 -1.01 -0.06 0.10 0.82 0.03 -0.04 -0.70	(3.36，3.29，-0.07) (3.47，3.83，0.36) (5.52，5.41，-0.11) (3.42，3.47，0.05) (1.84，1.82，-0.02) (2.24，2.44，0.20) (4.77，5.04，0.27) (3.38，2.71，-0.67) (4.80，4.58，-0.22) (2.01，1.73，-0.28) (1.92，1.89，-0.03) (2.06，2.12，0.06) (3.97，3.98，0.01) (2.06，1.95，-0.11) (2.40，2.38，-0.02) (3.83，3.66，-0.17) (3.59，5.49，1.90) (3.09，4.01，0.92) (2.17，2.31，0.14) (4.45，3.80，-0.65) (4.94，3.93，-1.01) (1.95，1.89，-0.06) (1.93，2.03，0.10) (2.24，3.06，0.82) (2.12，2.15，0.03) (1.79，1.75，-0.04) (3.46，2.76，-0.70)
					The whole city	2.81	2.89	0.08	(2.81，2.89，0.08)

Because in the calculating of the cotton rice period of crop-rotation and the cotton period of crop-rotation of rice, adopted different decipher areas, the former is the area of kind paddy rice next year that grows cotton then, and the latter is the area that last year, the kind paddy rice grew cotton then.If the paddy rice wheel is finished needed time y with cotton ₁Obviously greater than the cotton wheel being finished needed time x with paddy rice ₁So this just to replant behind the cotton time of continuous cropping than paddy rice longer the continuous cropping time of explanation after cotton replants paddy rice, the action of soaking that therefore can make full use of paddy rice more kills plants disease and pest remaining in the cotton soil, to reach better crop rotation effect, therefore for the comparisons of two polynary several periods of crop-rotation, at x ₁Under the identical situation, more to consider difference function value y ₁-x ₁Size, promptly to y ₁Value compares, y ₁The effect of the big more then crop rotation of value is good more, according to this conclusion, can carry out more deep analysis by his-and-hers watches 3, and can draw more useful results.

Claims (8)

1, the remote sensing estimation method of crop multielement crop rotation cycle, relate to agricultural and association area, the estimation object is with the polynary period of crop-rotation of counting the crops of describing in big geographical regional extent, at first select to cover the remote sensing image of the crops relevant with the estimation object, then decipher is carried out in the variation spatially of the crops in the remote sensing image, and with the data of decipher result as estimation polynary several periods of crop-rotation of crops, set up actual polynary several periods of crop-rotation that relevant formula and substitution interpret data are estimated crops again, this method has efficiently, simply, objective, characteristics such as be widely used, can be widely used in the estimation of all polynary several periods of crop-rotation.

2, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, relate to agricultural and association area, the estimation object is for being meant on same plot with the polynary period of crop-rotation of counting the crops of describing in big geographical regional extent, a kind of crop-planting is after the regular hour, plant the another kind of crop regular hour again, and the process that replaces down is called shift of crops always, the time that is spent is called the crop rotation time from a kind of crop to another kind of crop, be also referred to as the period of crop-rotation, the conclusion of shift of crops cycle studies is equally applicable to unit At All Other Times the research in shift of crops cycle with the time, on this basis, it is as follows that definition is used for describing polynary several periods of crop-rotation of big geographic area crop rotation

The period of crop-rotation between crops x and the y can describe with the polynary number in following m rank:

(x ₁，…，x _m，y ₁，…，y _m，f ₁(x ₁，…，x _m，y ₁，…，y _m)，…，f _p(x ₁，…，x _m，y ₁，…，y _m))

Wherein: x _i(i=1 ..., m) be the crops x that tried to achieve with m kind diverse ways and the period of crop-rotation of crops y; y _i(i=1 ..., m) be the crops y that tried to achieve with m kind diverse ways and the period of crop-rotation of crops x; f _j(x ₁..., x _m, y ₁..., y _m) (j=1 ..., p) be reflected in period of crop-rotation x for what ask with p kind diverse ways _iAnd y _i(i=1 ..., therefore m) the difference function between can be divided three classes the unit in polynary several periods of crop-rotation, and the first kind is relevant with the crop rotation of crops x and y, and second class is relevant with the crop rotation of crops y and x, and the 3rd class is relevant with the second class unit with the first kind,

Supposition now only is respectively x with the period of crop-rotation that the cotton period of crop-rotation of rice and cotton rice period of crop-rotation formula are tried to achieve between crops x and the y ₁And y ₁, x is described ₁And y ₁Between the cycle differentiation function be f ₁(x ₁, y ₁)=y ₁-x ₁, polynary several periods of crop-rotation of single order of then describing crops x and y crop rotation are (x ₁, y ₁, y ₁-x ₁), the crops x that describes with traditional single numerical value of usefulness compared with the period of crop-rotation between the y, and polynary several periods of crop-rotation of crops provide more comprehensively information, are a kind of describing methods of more science, and can obtain many important properties, with the polynary several period of crop-rotation (x of single order ₁, y ₁, y ₁-x ₁) be example: if y ₁-x ₁＞0, illustrate that then crop rotation makes the income of crops x bigger than the income of crop y, this is because of the crop rotation for crops x and y, if crop rotation is to make crop x have higher ecology and economic benefit, crops x and the length of the y period of crop-rotation have confidential relation with ecological and economic benefit usually so, and the shorter periods of crop-rotation and crops y and x of the period of crop-rotation of crops x and y are long, more favourable to crops x usually, promptly for the comparisons of two polynary several periods of crop-rotation, at x ₁Under the identical situation, more to consider difference function value y ₁-x ₁Size, promptly to y ₁Value compares, y ₁The effect of the big more then crop rotation of value is good more,

In sum, illustrate that as example the polynary several period of crop-rotation estimation equations of single order of crops are as follows with cotton and paddy rice wheel, but it is key character of the present invention that this formula is applicable to the estimation of all polynary several periods of crop-rotation of crops single order, can make similar analysis, polynary several periods of crop-rotation of the single order of crops x and y crop rotation for polynary several periods of crop-rotation of the m rank of crops:

(x ₁，y ₁，y ₁-x ₁)

Wherein:

3, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, it is characterized in that the described remote sensing image of at first selecting to cover the crops relevant with the estimation object is meant that it is continuous choosing per two width of cloth images on the time of remote sensing image for many years that are used for decipher, but on the time, can be interrupted between continuous per two width of cloth images.

4, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, it is characterized in that describedly then decipher being carried out in the variation spatially of the crops in the remote sensing image and being meant that served as basic combination to separate the data that translate from remote sensing image with adjacent 2 years, each combination comprise the previous year by the area of rotation crop and on identical plot the previous year replanted or take turns area by rotation crop at next year as another kind of crop.

5, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, it is characterized in that described and with the decipher result as the estimation polynary several periods of crop-rotation of crops data, setting up actual polynary several periods of crop-rotation that relevant formula and substitution interpret data estimate crops again is meant that crop rotation can describe with stationary stochastic process, have periodically and can calculate, the periodicity here is that all that calculate statistically participate in a kind of crop of crop rotations by needed year umber of another kind of crops shifting cultivation.

6, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, it is characterized in that described and with the decipher result as the estimation polynary several periods of crop-rotation of crops data, setting up actual polynary several periods of crop-rotation that relevant formula and substitution interpret data estimate crops again is meant two kinds of crops of first, second of participation crop rotation, first and second crop rotation and second and first crop rotation are two different notions, and their period of crop-rotation can be different.

7, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, it is characterized in that described and with the decipher result as the estimation polynary several periods of crop-rotation of crops data, setting up actual polynary several periods of crop-rotation that relevant formula and substitution interpret data estimate crops again is meant with respect to big geographic area, natural township or village (city or county) (or littler) are met the zone of independent sample requirement as the base unit that calculates the shift of crops cycle, the independent sample number that calculates the crop cycle will significantly be increased, minimizing is to the requirement of remote sensing image quantity, and can improve the precision in estimation shift of crops cycle.

8, the remote sensing estimation method of crop multielement crop rotation cycle according to claim 1, it is characterized in that described and with the decipher result as the estimation polynary several periods of crop-rotation of crops data, setting up actual polynary several periods of crop-rotation that relevant formula and substitution interpret data estimate crops again is meant that following is the described mathematical formulae of example with C crops and R crop rotation, derivation, result of calculation and application process are applicable to the polynary several periods of crop-rotation of all crops and the estimation of spatial variations

Suppose the interpret data of M the small towns N continuous year remote sensing satellite image that has obtained C crops and R crop rotation zone, so available following formula calculates C crops and R crop rotation cycle,

Wherein M is the small towns sum; NRA _IjBe that j small towns is in the area of i kind C crops, at the area of i+1 kind R crops; CCA _IjBe the area of j small towns i kind C crops,

In like manner, suppose the interpret data of M the small towns N continuous year remote sensing satellite image that has obtained R crops and C crop rotation zone, so available following formula calculates R crops and C crop rotation cycle,

Wherein M is the small towns sum; NCA _IjBe that j small towns is in the area of i-1 kind R crops, at the area of i kind C crops; CCA _IjBe the area of j small towns i kind C crops,

Supposition now only is respectively x with the period of crop-rotation that the cotton period of crop-rotation of above-mentioned rice and cotton rice period of crop-rotation formula are tried to achieve between crops x and the y ₁And y ₁, x is described ₁And y ₁Between the cycle differentiation function be f ₁(x ₁, y ₁)=y ₁-x ₁, polynary several periods of crop-rotation of single order of then describing crops x and y crop rotation are (x ₁, y ₁, y ₁-x ₁), the crops x that describes with traditional single numerical value of usefulness compared with the period of crop-rotation between the y, and polynary several periods of crop-rotation of crops provide more comprehensively information, are a kind of describing methods of more science, and can obtain many important properties, with the polynary several period of crop-rotation (x of single order ₁, y ₁, y ₁-x ₁) be example: if y ₁-x ₁＞0, illustrate that then crop rotation makes the income of crops x bigger than the income of crop y, this is because if for the crop rotation of crops x and y, crop rotation is to make crop x have higher ecology and economic benefit, crops x and the length of the y period of crop-rotation have confidential relation with ecological and economic benefit usually so, and the shorter periods of crop-rotation and crops y and x of the period of crop-rotation of crops x and y are long, more favourable to crops x usually, promptly for the comparisons of two polynary several periods of crop-rotation, at x ₁Under the identical situation, more to consider difference function value y ₁-x ₁Size, promptly to y ₁Value compares, y ₁The effect of the big more then crop rotation of value is good more, illustrate that as example the polynary several period of crop-rotation estimation equations of single order of crops are as follows with cotton and paddy rice wheel, but it is key character of the present invention that this formula is applicable to the estimation of all polynary several periods of crop-rotation of crops single order, can make similar analysis, polynary several periods of crop-rotation of the single order of crops x and y crop rotation for polynary several periods of crop-rotation of the m rank of crops:

(x ₁，y ₁，y ₁-x ₁)

Wherein: