3.1 Analysis of precipitation characteristics in dryland wheat.
The analyzing the precipitation and precipitation distribution of meteorological stations in the Loess Plateau from 1983 to 2020 (Table 1), it can be seen that the average precipitation in the rainfed wheat region in the Loess Plateau during fallow period is 296mm, and the range of precipitation during fallow period is 255-374 mm, and the coefficient of variation (CV) was 10%. The annual mean precipitation was 515 mm, the range of annual mean precipitation was 411-616 mm, and the coefficient of variation (CV) was 11%. The proportion of precipitation in leisure period to annual precipitation is 58% on average, and the proportion of precipitation in leisure period is between 51% and 65%, and the coefficient of variation is CV=6%.
Table.1
Location, annual precipitation and summer fallow precipitation of 19 national weather station in the dryland wheat
|
Year
|
Site
|
Longitude Latitude
|
Precipitation (mm)
|
TP
|
FP
|
FP/TP (%)
|
1981-2020
|
Houma
|
N: 35°37′ E:111°22′
|
495
|
268
|
54
|
1981-2020
|
Jincheng
|
N: 35°30′ E:112°51′
|
606
|
374
|
62
|
1983-2017
|
Linfen
|
N: 36°04′ E:111°34′
|
484
|
284
|
59
|
1981-2020
|
Xixian
|
N: 36°41′ E:110°56′
|
521
|
312
|
60
|
1981-2017
|
Xiangyuan
|
N: 36°32′ E:113°03′
|
460
|
268
|
58
|
1981-2017
|
Yangcheng
|
N: 35°29′ E:112°24′
|
602
|
343
|
57
|
1983-2020
|
Wenxi*
|
N: 35°10′ E:110°59′
|
487
|
253
|
52
|
1981-2020
|
Changzhi
|
N: 36°11′ E:113°06′
|
531
|
287
|
54
|
1981-2017
|
Xiji
|
N: 35°57′ E:105°43′
|
411
|
255
|
57
|
1981-2016
|
Guyuan
|
N: 36°0′ E:106°17′
|
452
|
266
|
59
|
1983-2015
|
Changwu*
|
N: 35°12′ E:107°47′
|
582
|
314
|
54
|
1981-2017
|
Qianxian*
|
N: 34°54′ E:108°58′
|
563
|
325
|
58
|
1981-2020
|
Yanan
|
N: 36°51′ E:109°19′
|
533
|
315
|
59
|
1981-2020
|
Luochuan
|
N: 35°45′ E:109°25′
|
616
|
342
|
55
|
1983-2020
|
Tongchuan
|
N: 35°04′ E:109°04′
|
587
|
319
|
54
|
1981-2015
|
Tianshui
|
N: 34°34′ E:105°53′
|
518
|
263
|
51
|
1981-2015
|
Yanglin*
|
N: 34°17′ E:108°05′
|
576
|
344
|
60
|
1981-2020
|
Huanxian
|
N: 36°34′ E:107°18′
|
434
|
259
|
60
|
1981-2020
|
Pingliang
|
N: 35°32′ E:106°40′
|
501
|
289
|
58
|
Mean
|
–
|
–
|
515
|
296
|
58
|
CV %
|
–
|
–
|
11
|
10
|
6
|
Note: The * represents sites Wenxi, Changwu, Qianxian and Yangling have corresponding long-term location experiments, can provide supporting literature yield data. TP indicate Annual precipitation, FP indicate Fallow season precipitation, CV indicate Variable coefficient. The same below
|
3.2 Correlation between precipitation and grain yield in dryland wheat region
Based on literature data (Table 2 and Figure 1), grain yield of dryland wheat increased logarithmically with the increase of precipitation in fallow season Figure 1A. With the increase of annual precipitation, grain yield of wheat in dryland increased logarithmically (Y = 3824LNX-19816.5, R2=0.45, P<0.01) Figure 1B.There was a non-linear and significant positive correlation between grain yield and precipitation in fallow period and annual precipitation in dryland wheat region.
Table 2
Annual precipitation, Fallow season precipitation and yield of dryland wheat from long term experiment in the Loess Plateau
|
Site
Literature
|
Year
|
TP (mm)
|
FP (mm)
|
Yield (kg ha–1)
|
Site
Literature
|
Year
|
TP (mm)
|
FP (mm)
|
Yield (kg ha–1)
|
Wenxi (Sun et al. 2018)
|
2009-2010
|
335
|
173
|
3924
|
Changwu (Guo et al. 2012)
|
1986-1987
|
503
|
222
|
2880
|
2010-2011
|
535
|
402
|
4795
|
1987-1988
|
481
|
203
|
3200
|
2011-2012
|
673
|
460
|
5412
|
1988-1989
|
574
|
234
|
2587
|
2012-2013
|
343
|
171
|
2915
|
1989-1990
|
778
|
510
|
5813
|
2013-2014
|
474
|
284
|
4819
|
1990-1991
|
550
|
277
|
4467
|
2014-2015
|
517
|
366
|
5000
|
1991-1992
|
668
|
395
|
3773
|
Yangling (Dai et al.2015)
|
2004-2005
|
573
|
386
|
6114
|
1992-1993
|
362
|
147
|
1080
|
2005-2006
|
570
|
452
|
4271
|
1993-1994
|
617
|
346
|
4067
|
2006-2007
|
354
|
179
|
5132
|
1994-1995
|
576
|
267
|
3640
|
2007-2008
|
618
|
459
|
3221
|
1995-1996
|
332
|
153
|
1120
|
2008-2009
|
491
|
310
|
5481
|
1996-1997
|
488
|
213
|
1960
|
2009-2010
|
695
|
422
|
6441
|
1997-1998
|
597
|
376
|
5893
|
2010-2011
|
593
|
402
|
6325
|
1998-1999
|
605
|
269
|
3440
|
2011-2012
|
880
|
680
|
6406
|
1999-2000
|
591
|
346
|
2920
|
2012-2013
|
549
|
387
|
3766
|
2000-2001
|
589
|
201
|
3120
|
Qiaxian 2002)
|
1987-1988
|
460
|
240
|
4125
|
2001-2002
|
427
|
152
|
6067
|
1988-1989
|
665
|
352
|
6158
|
2002-2003
|
675
|
314
|
4733
|
1989-1990
|
496
|
227
|
4526
|
2003-2004
|
442
|
185
|
3653
|
1990-1991
|
695
|
374
|
5595
|
2004-2005
|
532
|
310
|
5347
|
1991-1992
|
310
|
186
|
2079
|
2005-2006
|
525
|
299
|
3667
|
1992-1993
|
982
|
426
|
6518
|
2006-2007
|
509
|
333
|
3667
|
1993-1994
|
401
|
241
|
4307
|
2007-2008
|
598
|
344
|
4177
|
1994-1995
|
399
|
219
|
3290
|
2008-2009
|
568
|
315
|
3806
|
1995-1996
|
527
|
262
|
4133
|
Changwu (He et al. 2016)
|
2008-2009
|
513
|
270
|
3301
|
1996-1997
|
572
|
342
|
5984
|
2009-2010
|
475
|
280
|
3365
|
1997-1998
|
426
|
207
|
3231
|
2010-2011
|
666
|
458
|
5070
|
1998-1999
|
570
|
319
|
3650
|
2011-2012
|
722
|
453
|
7639
|
1999-2000
|
478
|
352
|
2978
|
2012-2013
|
447
|
285
|
3877
|
Note: TP indicate Annual precipitation, FP indicate Fallow season precipitation.
|
(Figure 1)
3.3 Annual classification of precipitation based on fallow period
This study refers to the classification method of precipitation year types proposed in the arid region of the Loess Plateau Taking Wenxi of Shanxi Province as an example, the 37 years from 1983 to 2020 are divided into dry years, perennial years and wet years, of which 7 years are wet years (annual precipitation >550 mm). 13 years were dry years (annual precipitation<440 mm), and 17 years were perennial years (440 mm< annual precipitation<550 mm) (Figure 2).Considering that the above classification method of precipitation year type is based on annual precipitation, it has no reference significance for farmers who need to make fertilization decision before sowing. Therefore, in order to establish the annual classification method based on precipitation in leisure period, this study further takes the average value of precipitation in leisure period from 1983 to 2020 in proportion to annual precipitation as the standard, and takes the abnormal years (abnormal years the dry years with too much precipitation in fallow period and the perennial years with too little precipitation in fallow period and the wet years) were eliminated, and the range of fallow precipitation in the remaining years was used as the basis for the annual classification of fallow precipitation. In all years, the years in which the proportion of precipitation in leisure period is higher than the annual average is defined as ordinary years (13 years in total, with a probability of 76.5%), and the years in which the proportion of precipitation in leisure period is lower than the annual average is defined as abnormal years (4 years in total, with a probability of 23.5%). In wet years, the years in which the proportion of precipitation in leisure period was higher than the multi-year average was defined as ordinary years (5 years in total, with a probability of 71.4%), and the years in which the proportion of precipitation in leisure period was lower than the multi-year average was defined as abnormal years (2 years in total, with a probability of 28.6%). In drought years, the years in which the proportion of precipitation in fallow period was lower than the multi-year average were defined as ordinary years (11 years in total, with a probability of 84.6%), and the years in which the proportion of precipitation in fallow period was higher than the multi-year average were defined as abnormal years (2 years in total, with a probability of 15.4%). Specific methods are as follows.
(Figure 2)
The probability of ordinary year for the three precipitation year types is more than 70% (perennial 76.5%, wet 71.4% and dry 84.6%) (Figure 3).Therefore, the precipitation range in fallow period of different precipitation year types can be determined by excluding abnormal years. Drought year, perennial year, and wet year were 85.5-220.7 mm (CV=26%), 262.5-439.3 mm (CV=16%), and 437.0-496.3 mm (CV=10%), respectively. Taking the upper limit of precipitation range in fallow season in dry years (<220 mm) and the lower limit of precipitation range in fallow season in wet years (>440 mm) as the thresholds for dividing year types based on fallow season precipitation, the results are as follows dry year (fallow precipitation <220 mm), perennial year (220 mm < fallow precipitation < 440 mm), wet year (fallow precipitation >440 mm).
(Figure 3)