Modeling wheat yield gaps in European Russia
using the SWAT model
SWAT conference 2013 | 17 – 19 July
Schierhorn, Florian Faramarzi, Monireh Prishchepov, Alexander V. Koch, Friedrich Müller, Daniel
guardian.co.uk
Rising demand for agricultural products
Population o 2.5 billion more by 2070
o Growing affluence
Diets o Meat
o Processed food
Bioenergy o Traditional (80%)
o Modern (20%)
guardian.co.uk
2 guardian.co.uk
Two options to increase agricultural production
1. Expansion of agricultural land o Most fertile land already used
o Further expansion associated with high environmental tradeoffs
2. Intensification on existing agricultural land o More inputs per unit of land
o Better technology
3
Low yields due to yield gap?
• Simulation of post-Soviet wheat yields using SWAT [->capturing of inter-anual yield variations]
• Simulation of wheat yield potentials
• Sensitivity analysis of growing period (& PHU) and crop parameters on yield
Objectives
#*
#* Kazakhstan
Kirov
Rostov
Saratov
Perm
Volgograd
Nizhny
Bashkortostan
Vologda
Tatarstan
Penza
Samara
KrasnodarStravropol
Tula
Moscow
Ryazan
Voronezh
Kursk
Oryol
Udmurtia
TambovBryansk
Ulyanovsk
MariEl
Lipetsk
Belgorod
Mordovia
Chuvash
70°E
60°E
60°E
50°E
50°E
40°E
40°E30°E
50°N
50°N
Ukraine
European Russia
Spring wheat
Subbasin
Provinces (oblast)
Study area
546 subbasins -> 30 selected for individual calibration
Model setup
Calibration: 1996-2006 / 3yrs warm-up
Validation: 1991-1994 / 3yrs warm-up
Study area size: 4 million km² square kilometers
Hargreaves method for ET-estimation
Dominant soil and land use property
Winter wheat & spring wheat dominant
Assignment of ONE representative subbasin per province (oblast)
-> 30 regions to be calibrated indi
Data
• Landuse: Binary cropland map (100 ha)
• Soil: Harmonized World Soil Database (10 km)
• Climate: Daily weather data from Schuol and Abbaspour (2007)
• Wheat yield: ROSSTAT 2012 (provinces)
• Management: • Growing period – USDA
• N fertilizer – ROSSTAT 2012
Calibration, Validation, and scenarios analysis
• Wheat yields were calibrated for each province
• Simulated wheat yields were compared with anual wheat yield data reported at provincial level
• After satisfactory calibration/validation, two scenarios: A) sufficient N fertilizer, B) sufficient N fertilizer + Water were constructed [to access yield potentials]
Initial parameter ranges for calibration
Parameter MIN MAX Parameter MIN MAX
v__ESCO.hru 0.01 1 v__USLE_K(1).sol 0 0.65
v__EPCO.hru 0.01 1 v__USLE_K(2).sol 0 0.65
v__OV_N.hru 0.01 0.8 v__SOL_EC(1).sol 0 100
v__LAT_TTIME.hru 0 180 v__SOL_EC(2).sol 0 100
v__LAT_SED.hru 0 5000 v__SHALLST.gw 0 1000
v__CANMX.hru 0 1 v__DEEPST.gw 0 1290
v__SOL_BD(1).sol 1.1 1.7 v__GW_DELAY.gw 0 500
v__SOL_BD(2).sol 1.1 1.7 v__ALPHA_BF.gw 0 1
v__SOL_CBN(1).sol 1 3 v__GWQMN.gw 0 5000
v__SOL_CBN(2).sol 0 2 v__GW_REVAP.gw 0.02 0.2
v__SOL_ALB(1).sol 0 0.25 v__RCHRG_DP.gw 0 1.1
v__SOL_ALB(2).sol 0 0.25 v__GWHT.gw 0 25
v__ANION_EXCL.sol 0.2 0.8 v__GW_SPYLD.gw 0 0.4
v__SOL_K(1).sol 0 2000 v__SHALLST_N.gw 0 10
v__SOL_K(2).sol 0 2000 v__GWSOLP.gw 0 1000
v__SOL_CRK.sol 0 1 v__HLIFE_NGW.gw 0 365
v__CN2.mgt 65 90 r__FRT_KG{x,x}.mgt -0.3 0.3
Crop related parameters were excluded for calibration!
Parameter 129 139 201 220 221 235 251 279 300 308 379 425
FRT_KG{12,3}.mgt 8 9 14 10 8 7 5 11 18 1 5 1
SOL_CBN(2).sol 7 1 8 26 1 8 6 12 1 6 12 9
FRT_KG{9,3}.mgt 20 2 3 1 15 15 19 1 20 2 1 3
FRT_KG{6,3}.mgt 9 7 48 4 5 11 1 3 11 4 6 15
FRT_KG{11,3}.mgt 5 15 21 9 6 9 2 8 27 3 13 11
FRT_KG{14,3}.mgt 10 12 15 12 2 1 10 2 23 12 11 20
FRT_KG{8,3}.mgt 29 3 1 6 16 14 3 5 31 16 2 4
ESCO.hru 6 13 4 3 14 4 16 41 2 23 3 2
FRT_KG{10,3}.mgt 3 10 6 8 4 34 8 7 10 13 25 5
CN2.mgt 48 5 2 2 13 6 14 10 5 14 9 8
FRT_KG{7,3}.mgt 2 4 5 5 19 12 4 4 24 43 8 6
FRT_KG{13,3}.mgt 11 16 12 11 7 3 15 6 35 5 10 16
SOL_CBN(1).sol 12 6 11 28 12 10 12 13 3 9 23 13
FRT_KG{15,3}.mgt 1 17 23 7 3 18 7 14 47 8 7 17
FRT_KG{4,3}.mgt 4 11 9 14 21 25 9 37 28 10 4 12
ANION_EXCL.sol 26 8 10 19 10 5 13 25 8 15 29 18
EPCO.hru 35 20 7 13 11 2 18 15 25 7 17 19
FRT_KG{5,3}.mgt 44 23 13 16 9 21 11 9 9 36 14 7
CANMX.hru 15 24 16 15 17 13 29 46 19 17 16 44
SOL_BD(1).sol 16 14 43 35 18 40 17 28 6 21 21 36
SOL_BD(2).sol 14 18 32 18 47 16 47 23 4 32 30 14
SOL_K(1).sol 42 34 26 41 29 28 23 27 7 11 18 10
GWSOLP.gw 23 32 28 17 27 23 24 32 12 22 19 43
FRT_KG{3,3}.mgt 25 19 27 23 37 24 31 18 30 44 24 23
GWQMN.gw 18 29 24 45 30 20 38 30 16 18 33 34
SOL_K(2).sol 39 33 22 30 31 42 26 16 41 19 40 27
LAT_SED.hru 37 30 33 36 28 37 20 20 17 47 42 22
SHALLST_N.gw 13 40 40 20 25 19 22 39 43 46 41 24
FRT_KG{1,3}.mgt 28 31 25 47 24 38 39 42 26 20 31 26
SOL_EC(1).sol 24 22 38 43 40 26 27 22 36 48 26 25
DEEPST.gw 40 25 44 37 20 32 42 19 32 45 15 30
GWHT.gw 30 37 29 24 33 30 44 17 46 29 45 21
LAT_TTIME.hru 22 36 46 32 46 17 46 38 14 26 32 46
FRT_KG{2,3}.mgt 21 46 31 27 42 33 25 24 48 24 37 45
SOL_ALB(1).sol 19 43 18 40 43 36 40 40 15 38 43 29
SOL_EC(2).sol 36 26 30 46 34 22 35 21 33 37 47 37
GW_REVAP.gw 46 27 34 31 41 46 21 26 40 27 38 31
HLIFE_NGW.gw 45 35 20 21 45 31 32 45 38 30 35 38
OV_N.hru 41 48 39 22 26 48 45 29 13 35 44 28
GW_SPYLD.gw 38 42 36 33 22 47 41 35 21 40 28 40
SOL_ALB(2).sol 17 41 42 39 35 29 43 47 45 28 22 35
USLE_K(1).sol 47 28 47 25 38 27 30 43 34 34 36 41
RCHRG_DP.gw 31 47 41 42 23 35 33 48 29 31 34 39
ALPHA_BF.gw 43 44 35 38 39 39 34 31 39 42 20 33
USLE_K(2).sol 27 21 37 29 44 45 48 36 44 39 27 42
SHALLST.gw 32 38 19 34 48 41 37 33 42 33 39 48
SOL_CRK.sol 34 39 45 44 36 43 28 34 22 25 48 47
GW_DELAY.gw 33 45 17 48 32 44 36 44 37 41 46 32
Param
eter se
nsitivity
+
-
Calibration and Validation results
Calibration Validation
0
500
1000
1500
2000
2500
3000
3500
0
0.5
1
1.5
2
1 10 19 28 37 46 55 64 73 82 91 100 109
10 kg NNo irrigation
PHU: 2009 YLD = 0.7 t/ha
0
1000
2000
3000
4000
5000
6000
7000
8000
0
0.5
1
1.5
2
2.5
3
3.5
4
1 11 21 31 41 51 61 71 81 91 101 111
PHU: 2009 YLD = 1.2 t/ha
300 kg NNo irrigation
0
2000
4000
6000
8000
10000
12000
14000
16000
0
0.5
1
1.5
2
2.5
3
3.5
4
1 11 21 31 41 51 61 71 81 91 101 111
PHU: 2009 YLD = 4.7 t/ha
300 kg NIrrigation
0
2000
4000
6000
8000
10000
12000
0
0.5
1
1.5
2
2.5
3
3.5
4
1 11 21 31 41 51 61 71
300 kg NIrrigation
PHU: 1409 YLD = 3.4 t/ha
Rainfed Rainfed
02468
02468
02468
02468
02468
1995 2000 2005 1995 2000 2005
1995 2000 2005 1995 2000 2005 1995 2000 2005 1995 2000 2005
Bashkortostan Belgorod Bryansk Chuvash Kirov Krasnodar
Kursk Lipetsk MariEl Mordovia Moscow Nizhny
Oryol Penza Perm Rostov Ryazan Samara
Saratov Stravropol Tambov Tatarstan Tula Udmurtia
Ulyanovsk Volgograd Vologda Voronezh
Yield potential for sufficient N & WATER supply
Yield potential for sufficient N supply
Reported
Wh
eat
yiel
d (
t/h
a)
0
1
2
3
4
5
6
7
Wh
eat
yie
ld (
t/h
a)
YGN
YGN&Water
Yield gap for sufficient N & Water supply
#*
#*
Kirov
Rostov
Saratov
Perm
Volgograd
Nizhny
Bashkortostan
Vologda
Tatarstan
Penza
Samara
KrasnodarStravropol
Tula
Moscow
Ryazan
Voronezh
Kursk
Oryol
Udmurtia
TambovBryansk
Ulyanovsk
MariEl
Lipetsk
Belgorod
Mordovia
Chuvash
70°E
60°E
60°E
50°E
50°E
40°E
40°E30°E
50°N
50°N
Ukraine
European Russia
t/ha
1.10 - 1.18
1.19 - 1.35
1.36 - 1.48
1.49 - 1.87
1.88 - 2.16
2.17 - 2.43
2.44 - 2.53
2.54 - 2.94
2.95 - 3.15
3.16 - 4.31
Average wheat yieldsreported 2001-2006
Spring wheat
Average wheat yields, 2001-2006
#*
#*
Kirov
Rostov
Saratov
Perm
Volgograd
Nizhny
Bashkortostan
Vologda
Tatarstan
Penza
Samara
KrasnodarStravropol
Tula
Moscow
Ryazan
Voronezh
Kursk
Oryol
Udmurtia
TambovBryansk
Ulyanovsk
MariEl
Lipetsk
Belgorod
Mordovia
Chuvash
70°E
60°E
60°E
50°E
50°E
40°E
40°E30°E
50°N
50°N
Ukraine
European Russia
t/ha
1.16 - 1.43
1.44 - 1.75
1.76 - 1.80
1.81 - 1.97
1.98 - 2.17
2.18 - 2.31
2.32 - 2.37
2.38 - 2.52
2.53 - 2.82
2.83 - 3.03
Average yield gap for sufficient N supply, 2001-2006
Spring wheat
Yield gap for sufficent N supply
Yield gap for sufficent N supply & water supply
#*
#*
Kirov
Rostov
Saratov
Perm
Volgograd
Nizhny
Bashkortostan
Vologda
Tatarstan
Penza
Samara
KrasnodarStravropol
Tula
Moscow
Ryazan
Voronezh
Kursk
Oryol
Udmurtia
TambovBryansk
Ulyanovsk
MariEl
Lipetsk
Belgorod
Mordovia
Chuvash
70°E
60°E
60°E
50°E
50°E
40°E
40°E30°E
50°N
50°N
Ukraine
European Russia
t/ha
1.79 - 2.57
2.58 - 2.74
2.75 - 2.89
2.90 - 3.11
3.12 - 3.20
3.21 - 3.38
3.39 - 3.44
3.45 - 3.72
3.73 - 3.80
3.81 - 4.56
Average yield gap for sufficient N & Water supply, 2001-2006
Spring wheat
Yield gap for entire European Russia: 2.1 t/ha -> under sufficient N supply 3.2 t/ha -> under sufficient N & Water supply
34567
34567
34567
34567
1995 2000 2005
1995 2000 2005 1995 2000 2005 1995 2000 2005
Bashkortostan Chuvash Kirov MariEl
Mordovia Nizhny Penza Perm
Samara Saratov Tatarstan Udmurtia
Ulyanovsk Volgograd Vologda
95PPU of simulated yield, average growing period
95PPU of simulated yield, short growing period
95PPU of simulated yield, long growing period
Spri
ng
wh
eat
yiel
d (
t/h
a)
Yield sensitivity to growing period
02468
02468
02468
1995 2000 2005
1995 2000 2005 1995 2000 2005
Mordovia Nizhny Penza
Samara Saratov Tatarstan
Ulyanovsk Volgograd
95PPU of simulated yield for sufficient N & Water supply, Variation of crop parameters
95PPU of simulated yield for sufficient N & Water supply, SWAT default crop parameters
Reported
Spri
ng
wh
eat
yiel
d (
t/h
a)
Yield sensitivity to crop parameter
Conclusion
• Russia has large potential to increase crop yields
• SWAT crop growth module reproduces accurate yields despite the large scale & data scarcity
• Yield gaps are heterogeneously distributed
• Information on crop characteristics (PHU in particular) are essential for accurate yield gap estimations
20
-.5 0 .5 1 1.5 2 -.5 0 .5 1 1.5 2
-.5 0 .5 1 1.5 2 -.5 0 .5 1 1.5 2
Validation
Calibration
Validation
Calibration
Validation
Calibration
Validation
Calibration
Bashkortostan Chuvash Kirov MariEl
Mordovia Nizhny Penza Perm
Samara Saratov Tatarstan Udmurtia
Volgograd Vologda P factor
R factor
R2
Nash Sutclif
Calibration and Validation results