Climate and Irrigation Management Scenarios

sumanager56 commented 1 year ago

Scenario_0: baseline

Rplot

Scenario_1: warming only (+2.75 deg C)

Rplot

Scenario_2: warming (+2.75 deg C) plus greater precip (+15%)

Rplot

Scenario_3: warming (+2.75 deg C) plus less precip (-5%)

Rplot

Scenario_4: warming (+2.75 deg C) plus increasingly variably rainfall (gini increases by 0.04%)

Rplot

Scenario_5: warming (+2.75 deg C), greater precip (+15%), and greater gini (+0.04%)

Rplot

sumanager56 commented 1 year ago

Updated Calendar-based irrigation with increased crop water use with the increase in temperature +2.4 degrees C

Scenario_0: baseline

Rplot

Scenario_1: warming only (+2.75 deg C)

Rplot_newET

Scenario_2: warming (+2.75 deg C) plus greater precip (+15%)

NewET

Scenario_3: warming (+2.75 deg C) plus less precip (-5%)

Rplot_newET

Scenario_4: warming (+2.75 deg C) plus increasingly variably rainfall (gini increases by 0.04%)

Rplot_newET

Scenario_5: warming (+2.75 deg C), greater precip (+15%), and greater gini (+0.04%)

Rplot_newET

sumanager56 commented 1 year ago

Rainfed vs Conventional vs Precision Irrigation and Nutrient Management Scenarios

Rainfed: No Irrigation/100% N application at planting Conventional: Weekly ET requirement minus Total Weekly Rainfall/ 100% N application at planting Precision: Irrigated at 50% of total available water content up to FC minus 15% of Available water content/25% N applied at planting and 75% on June 5th

Scenario_0: baseline

Rplot

Scenario_1: warming only (+2.75 deg C)

Rplot

Scenario_2: warming (+2.75 deg C) plus greater precip (+15%)

Rplot

Scenario_3: warming (+2.75 deg C) plus less precip (-5%)

Rplot

Scenario_4: warming (+2.75 deg C) plus increasingly variably rainfall (gini increases by 0.04%)

Rplot

Scenario_5: warming (+2.75 deg C), greater precip (+15%), and greater gini (+0.04%)

Rplot

sumanager56 commented 1 year ago

@julieshortridge Dr. Shortridge, Since we are updating the calendar-based irrigation for different climate scenarios, I wanted to see if updating the Precision irrigation thresholds would improve the results. Setting the threshold to 70% depletion instead of 50% depletion increased the corn and Nitrogen yield and reduced the Nitrate leaching as compared to the calendar-based treatment and 50% threshold-based results.

Summary of Mean Values

Parameters | Treatments | Scenario_0 | Scenario_1 | Scenario_2 | Scenario_3 | Scenario_4 | Scenario_5 -- | -- | -- | -- | -- | -- | -- | -- Corn_Yield | Rainfed | 170.926487 | 119.7752 | 133.0051 | 113.3711 | 117.203 | 129.5902 | Calendar | 202.0060596 | 175.488 | 170.1188 | 178.6932 | 175.2783 | 168.7075 | Precision | 222.647903 | 181.082 | 176.2948 | 182.4358 | 180.0462 | 174.6356 N-Yield | Rainfed | 166.8717644 | 117.1104 | 124.9376 | 113.2878 | 114.8989 | 122.7673 | Calendar | 173.8174498 | 149.2966 | 144.3526 | 152.3376 | 149.9008 | 143.1173 | Precision | 195.4168858 | 154.5989 | 150.7277 | 155.6795 | 154.1918 | 149.8484 N-Leach | Rainfed | 92.52550234 | 163.7745 | 157.4882 | 166.7938 | 165.101 | 159.056 | Calendar | 85.88195277 | 138.4339 | 140.3013 | 137.7319 | 139.7376 | 142.015 | Precision | 70.79376321 | 137.9746 | 139.65 | 137.5918 | 138.5941 | 141.0355

Scenario-0

Rplot

Scenario-1

Rplot

Scenario-2

Rplot

Scenario-3

Rplot

Scenario-4

Rplot

Scenario-5

Rplot

julieshortridge commented 1 year ago

Ok, fantastic! I think that's a good change. Looking at the summary table at the top of your post, I think my main question is what is driving the big drop off in yields from Scenario 0 (historic) to Scenario 1 (increased T only, right?). Do you have a sense of what is driving that? I don't necessarily think it's wrong, as there is some evidence suggesting that higher temperatures on their own will depress crop yields here in Virginia since it's already hot here, but I wouldn't expect that drastic of a decline. So I'm wondering if it stems from a mismatch in growing conditions - maybe there's less time to maturity, but water needs are greater over that time period?

Once we have an answer for that, I think the next step should be to show Dr. Stewart a summary of these results and see if he has any final suggestions or comments, then we can pass them along to Dr. Zhang.

Good work!

sumanager56 commented 1 year ago

Hi, Dr. Shortridge @julieshortridge . I checked the reference temperature (Temp_ref) parameter that controls the transformation rates of nitrification process (It can be used as a part of calibration process or the mean annual air temperature can be used as the default value). I had set this parameter to 15 degrees before, but this time I used the average temperature through the growing season (22 degrees C) and that significantly reduced the leaching values. Increasing this ref temperature reduces leaching but marginally change the yield and nitrogen uptake values. I believe we have more reasonable leaching values using the average growing season temperature (April-September).

Also, for Scenario1 through Scenerio5, I set it to 25 which is again the average that I calculated from the weather data. Besides, I set the planting date to April 1 for Scenario1-Scenario5, which provided reasonable yield estimates.

For Precision treatment, Nitrogen application is done 25% on the planting date and 75% on 16th May (Scenario 1-5)/05th June (Scenario 0) (both at the same development stage; June 5 was selected from Lal's paper). Calendar and Rainfed treatments are fertilized at the planting date only.

Precision irrigation is triggered at 50% this time and applied until 85% of the FC.

Summary of Mean Values

Parameters | Treatments | Scenario_0 | Scenario_1 | Scenario_2 | Scenario_3 | Scenario_4 | Scenario_5 -- | -- | -- | -- | -- | -- | -- | -- Corn_Yield | Rainfed | 165 | 128 | 145 | 120 | 125 | 142 | Calendar | 186 | 191 | 187 | 192 | 192 | 186 | Precision | 218 | 197 | 194 | 198 | 196 | 194 N-Yield | Rainfed | 159 | 144 | 152 | 137 | 141 | 149 | Calendar | 163 | 177 | 172 | 177 | 181 | 172 | Precision | 187 | 184 | 181 | 184 | 183 | 181 N-Leach | Rainfed | 69 | 114 | 108 | 118 | 117 | 110 | Calendar | 67 | 91 | 93 | 91 | 91 | 94 | Precision | 49 | 89 | 90 | 89 | 89 | 91

Scenario-0

Rplot

Scenario-1

Rplot

Scenario-2

Rplot

Scenario-3

Rplot

Scenario-4

Rplot

Scenario-5

Rplot

sumanager56 commented 1 year ago

Dr. @julieshortridge and Dr. @ryestewart, Here's some updates based on our discussion on last week's meeting. 1) Temp_ref: I think I should keep this parameter value as it is default:15 degree C and do not change under different climate scenarios because our rate coefficients like (De)Nitrification rates are used/calibrated for this same temperature and we are not going to change these rates in future scenarios. I found a paper that calibrated this coefficient to 7.5 degree C for their location since they had a mean annual temperature of around 9 degree C. For Suffolk, 15 degree C should be good. Here's also the plot showing lower response rates with high reference temperature. At same temperature, there is higher rate of mineralization, and so the nitrification and leaching values if the reference temperature is less. 15degrees 22degrees 2) Denitrification: Upon checking the denitrification values in my previous N balance table, the total denitrification for 20 years was just 6 kg/ha. This is far too low-studies suggest this value ranging from 10-50 kg/ha/year (200-1000 kg/ha in 20 years). So, I went back to see a parameter WFPScrit2 which is the critical water filled pore space value for the denitrification rate. The default value is 0.7 but I changed it to 0.95 previously because I was observing very high denitrification rate previously when some other parameters probably were not set well. A value of 0.95 means that denitrification only occurs when soil is close to saturation (>95% pore space filled with water) which is not reasonable. This store NO3 too much in the soil and also is a cause of huge NO3 leaching. Setting back this parameter to default 0.7 value provided reasonable denitrification rates of around 10 kg/ha/year. This also reduced leaching with yield and N uptake values only changed slightly. Here's the current balance for precision irrigation treatment (total of 20 years). <html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1800 | Volatilisation | 270 | Ammendments | 1800 | | Crop Residues | 0 | Plant uptake | 1866 | Crop Residues | 0 | Plant uptake | 1785 Mineralisation | 1486 | Nitrification | 866 | Nitrification | 866 | Denitrification | 186 Influx top | 30 | Outflux | 313 | Influx top | 30 | Outflux | 721 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3316 | Total | 3315 | Total | 2696 | Total | 2692 | | Diff.Stor | 1 | | | Diff.Stor | 4

3) Also, the initial organic matter is somewhere between 1-1.5%. 4) Seasonal variation in leaching

4) The NO3 (30 kg/ha/yr) and NH4 influx (30 kg/ha/yr) is based on default SWAP conditions but this seems pretty high for the US as per literature. Upon confirming from multiple studies, average of 3.2kg/ha/yr is the annual wet deposition of N in the US (average of 226 sites studied in 2017) (1.6kg/ha/yr each NO3 and NH4). I updated the influx rate and this also reduced leaching values.

Results showing mean of 20 years

Parameters | Treatments | Scenario_0 | Scenario_1 -- | -- | -- | -- Corn_Yield (bu/acre) | Rainfed | 159 | 119 | Calendar | 157 | 171 | Precision | 213 | 178 N-Yield (kg/ha) | Rainfed | 155 | 117 | Calendar | 147 | 142 | Precision | 182 | 151 N-Leach (kg/ha) | Rainfed | 60 | 105 | Calendar | 54 | 73 | Precision | 36 | 77

sumanager56 commented 1 year ago

Rainfed: No Irrigation/100% N application at planting Conventional: Weekly ET requirement minus Total Weekly Rainfall/ 100% N application at planting Precision: Irrigated at 50% of total available water content up to FC minus 15% of Available water content/25% N applied at planting and 75% on June 5th Planting dates-April 20th for all scenarios

Parameters | Treatments | Scenario_0 | Scenario_1 | Scenario_2 | Scenario_3 | Scenario_4 | Scenario_5 -- | -- | -- | -- | -- | -- | -- | -- Corn_Yield (bu/acre) | Rainfed | 159 | 120 | 133 | 113 | 116 | 128 | Calendar | 157 | 172 | 164 | 174 | 171 | 164 | Precision | 214 | 178 | 175 | 180 | 178 | 173 N-Yield (kg/ha) | Rainfed | 164 | 117 | 124 | 113 | 114 | 121 | Calendar | 166 | 147 | 135 | 144 | 142 | 135 | Precision | 194 | 153 | 149 | 152 | 151 | 147 N-Leach (kg/ha) | Rainfed | 60 | 105 | 96 | 110 | 108 | 100 | Calendar | 54 | 73 | 74 | 73 | 75 | 75 | Precision | 36 | 78 | 76 | 78 | 79 | 78

Scenario-0

Rplot

Scenario-1

Rplot

Scenario-2

Rplot

Scenario-3

Rplot

Scenario-4

Rplot

Scenario-5

Rplot

sumanager56 commented 1 year ago

The NO3 leaching in the above scenarios are generally higher for the precision treatment compared to the calendar-based irrigation although yields and N uptake are higher for the precision. However, by setting the FC lower towards 180 cm (FC was set to 250 cm in all previous scenarios) and triggering irrigation at 30% depletion would provide more frequent and higher total irrigation to the crops and the leaching is reduced significantly in precision treatment as compared to calendar-based treatment. Here's an example plot for Scenario-1 just by setting FC to 180 cm and maximum allowed depletion to 30% of PAWC.

Scenario-1

Irrthres-0 7FC180

ryestewart commented 1 year ago

What is the depth over which FC is calculated? What is the equivalent value in cm/cm or cm3/cm3?

Thanks, Ryan

On Sun, Aug 13, 2023 at 4:33 PM Suman Budhathoki @.***> wrote:

The NO3 leaching in the above scenarios are generally higher for the precision treatment compared to the calendar-based irrigation although yields and N uptake are higher for the precision. However, by setting the FC lower towards 180 cm (FC was set to 250 cm in all previous scenarios) and triggering irrigation at 30% depletion would provide more frequent and higher total irrigation to the crops and the leaching is reduced significantly in precision treatment as compared to calendar-based treatment. Here's an example plot for Scenario-1 just by setting FC to 180 cm and maximum allowed depletion to 30% of PAWC. Scenario-1

[image: Irrthres-0 7FC180] https://user-images.githubusercontent.com/99036349/260325759-936ed308-9f5d-4358-b421-df953354d537.png

— Reply to this email directly, view it on GitHub https://github.com/sumanager56/SWAP-WOFOST/issues/15#issuecomment-1676456297, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3TXL5TFGDDLPMDOX35H4T3XVE2XZANCNFSM6AAAAAAYMJXBYU . You are receiving this because you were mentioned.Message ID: @.***>

sumanager56 commented 1 year ago

Thanks, Dr. Stewart @ryestewart . SWAP calculates the water content corresponding to different head values independently for each compartment within the root zone (determines lowest compartment containing roots first at a given development stage). It uses pressure head at FC (-180 cm), h below which there's no water extraction (-7000 cm), h below which water uptake reduction starts (-350 cm) for each compartment to calculate the water holding capacity, readily available water, actual available water and water deficit, and then sums it for the entire root zone and schedules irrigation based on that. However, it does not provide a value for water content at FC- though I can figure out the corresponding water content at FC for different soil layers using the vg function. Please let me know if that answers your question.

julieshortridge commented 1 year ago

Looking at the table you posted before you changed the FC and irrigation threshold values (the comment that starts with the description of the management treatments), most of that table is as I would expect. Precision irrigation results in higher yields and higher N uptake than the calendar treatment. While those increases are fairly small, I think that's ok. The strange result is that it ALSO results in higher N leaching. So precision irrigation means 1) less water applied as expected, 2) higher N uptake as expected since yields are higher, and 3) higher N leaching which is unexpected.

So instead of changing the precision irrigation rules, I would try to understand WHY that treatment is leading to higher N leaching. So I think you can compare the N balance for precision versus calendar treatment (just do scenario 1 for now) to get a sense of that.

Also, wouldn't changing the FC values also change your soil moisture through time since that would lead to greater (or less) water draining out the bottom of the root zone for a given amount of rainfall? In other words, wouldn't it change things beyond just the irrigation response? If that's the case, then I think we definitely don't want to be changing it at this point. But maybe I'm misunderstanding the role it plays in determining soil moisture dynamics in the model.

sumanager56 commented 1 year ago

Thank you Dr. Shortridge, @julieshortridge . Looking into the N balance for precision versus calendar treatment (Scenario 1), it shows that although plant uptake is higher in precision treatment, most of them are in the form of NH4. NO3 uptake on the other hand, are lesser for precision, leading to more NO3 losses. Also, denitrification rates (exists with water filled pore space>70%) are lesses in the precision treatment which is as expected, again creating chances for more NO3 losses in the precision treatment. Although the ammendments are same for both the treatments, there is much higher mineralization rates in the precision treatment because of relatively drier conditions compared to the calendar-based (transformation rates is reduced when water filled pore space>95% which is more likely in calendar-based treatment). When I look into the overall leaching component (NH4+NO3), the leaching values are almost similar in both the treatments.

Scenario-1 Calendar

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1800 | Volatilisation | 270 | Ammendments | 1800 | | Crop Residues | 0 | Plant uptake | 1775 | Crop Residues | 0 | Plant uptake | 1069 Mineralisation | 2081 | Nitrification | 1325 | Nitrification | 1325 | Denitrification | 617 Influx top | 33 | Outflux | 534 | Influx top | 33 | Outflux | 1463 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3914 | Total | 3904 | Total | 3158 | Total | 3149 | | Diff.Stor | 10 | | | Diff.Stor | 9

Scenario-1 Precision

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1800 | Volatilisation | 270 | Ammendments | 1800 | | Crop Residues | 0 | Plant uptake | 2046 | Crop Residues | 0 | Plant uptake | 980 Mineralisation | 2249 | Nitrification | 1288 | Nitrification | 1288 | Denitrification | 563 Influx top | 31 | Outflux | 465 | Influx top | 31 | Outflux | 1558 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 4080 | Total | 4069 | Total | 3119 | Total | 3101 | | Diff.Stor | 11 | | | Diff.Stor | 18

Changing the FC values do not change the soil moisture dynamics in the SWAP model. I think this is one of the benefit of using a Richard's equation based model (like SWAP and Hydrus) that describes the change in volumetric water content over time by considering complex processes like hydraulic conductivity, pressure gradients, and root water extraction.
Different other models uses the concept of field capacity, meaning each soil layer has a specific water storage capacity and when soil water content exceeds FC, gravitational drainage occurs. They are also called the soil water balance or bucket-type models.

sumanager56 commented 1 year ago

Also, looking into the corresponding VWC values at -250 cm pressure head (FC) for the different depths, here's the results. 1) Layer 0-15 cm - 19.5% 2) Layer 15-40 cm - 24.7% 3) Layer 40-55 cm - 30.4% 4) Layer 55-85 cm - 35% 5) Layer 85-103 cm - 32.9% 6) Layer 103-125 cm - 52.4%

julieshortridge commented 1 year ago

Thanks Suman @ryestewart do the differences in nitrogen balance across the two irrigation treatments make sense to you as a reason you'd see higher nitrate leaching in the precision treatment rather than the full? Also, since we describe plant uptake in terms of total N, maybe we should describe N leaching in terms of total N (not just nitrate)?

sumanager56 commented 1 year ago

@julieshortridge Dr. Shortridge, I have updated the above results with the total N-leach(NH4+NO3). Also, here are the new plots showing total N leaching. <html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

Parameters | Treatments | Scenario_0 | Scenario_1 | Scenario_2 | Scenario_3 | Scenario_4 | Scenario_5 -- | -- | -- | -- | -- | -- | -- | -- Corn_Yield (bu/acre) | Rainfed | 159 | 119 | 133 | 113 | 116 | 128 | Calendar | 157 | 172 | 164 | 174 | 171 | 164 | Precision | 213 | 178 | 174 | 181 | 178 | 174 N-Yield (kg/ha) | Rainfed | 155 | 117 | 123 | 113 | 114 | 121 | Calendar | 147 | 142 | 135 | 142 | 142 | 135 | Precision | 182 | 151 | 148 | 150 | 151 | 147 N-Leach (kg/ha) | Rainfed | 79 | 132 | 126 | 135 | 135 | 129 | Calendar | 75 | 100 | 104 | 99 | 101 | 105 | Precision | 52 | 101 | 103 | 100 | 102 | 104

Scenario-0

Rplot

Scenario-1

Rplot

Scenario-2

Rplot

Scenario-3

Rplot

Scenario-4

Rplot

Scenario-5

Rplot

ryestewart commented 1 year ago

Hi all,

I'm still a bit worried that the nitrogen leaching values are pretty large

from what I'm seeing it seems like 20-35 kg N/ha is a more typical range for coastal plain soils. We might want to check in with Hunter and see if he has any sense of what a reasonable value would be.

I also think the field capacity water contents seems pretty high at depth, particularly the lowest layer. What is the porosity for that layer? My guess would be 40-45%, yet the field capacity is 52%.

Thanks, Ryan

Ryan D. Stewart, Ph.D. Associate Professor School of Plant and Environmental Sciences Virginia Tech, 241 Smyth Hall, Blacksburg, VA 24061 Phone: (540) 231-0253; Email: @. @.> Zoom: https://virginiatech.zoom.us/my/ryanstewart

On Wed, Aug 23, 2023 at 8:42 PM Suman Budhathoki @.***> wrote:

@julieshortridge https://github.com/julieshortridge Dr. Shortridge, I have updated the above results with the total N-leach(NH4+NO3). Also, here are the new plots showing total N leaching.

Parameters Treatments Scenario_0 Scenario_1 Scenario_2 Scenario_3 Scenario_4 Scenario_5 Corn_Yield (bu/acre) Rainfed 159 119 133 113 116 128 Calendar 157 172 164 174 171 164 Precision 213 178 174 181 178 174 N-Yield (kg/ha) Rainfed 155 117 123 113 114 121 Calendar 147 142 135 142 142 135 Precision 182 151 148 150 151 147 N-Leach (kg/ha) Rainfed 79 132 126 135 135 129 Calendar 75 100 104 99 101 105 Precision 52 101 103 100 102 104 Scenario-0

[image: Rplot] https://user-images.githubusercontent.com/99036349/262828240-e4dac564-865e-44a5-a238-dcf2f19c6e7f.png Scenario-1

[image: Rplot] https://user-images.githubusercontent.com/99036349/262828268-ebe22bc0-6c8b-40ad-8d07-675d0702a51e.png Scenario-2

[image: Rplot] https://user-images.githubusercontent.com/99036349/262828318-0b06de85-1670-4a6b-bc47-2ca378ad6a5d.png Scenario-3

[image: Rplot] https://user-images.githubusercontent.com/99036349/262828350-f48608b6-349c-4402-83a8-c4a3280824cc.png Scenario-4

[image: Rplot] https://user-images.githubusercontent.com/99036349/262828363-a2225778-106e-432f-a0ec-7366a3556d3b.png Scenario-5

[image: Rplot] https://user-images.githubusercontent.com/99036349/262828392-827a00b8-f6fb-4802-a17b-188fe23ae423.png

— Reply to this email directly, view it on GitHub https://github.com/sumanager56/SWAP-WOFOST/issues/15#issuecomment-1690820566, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3TXL5RAZPCEDDWJTNJPTBLXW2PNVANCNFSM6AAAAAAYMJXBYU . You are receiving this because you were mentioned.Message ID: @.***>

julieshortridge commented 1 year ago

Thanks Ryan. Suman, I was looking back at the N balances from when we were doing the single-year simulations (the issue "Nitrogen balance and corn yield compared to observed data") and it looks like the N leaching was high then as well - 67 kg/ha for the 0N treatment and 100 kg/ha for the 180N treatment. At that point, I know we were mostly focusing on comparing the N uptake values with the observed N uptake and that was reasonable (60 kg/ha uptake in 0N, 187 kg/ha uptake in 180N). I wonder if there is some other element of the N balance that is contributing to the high leaching numbers?

sumanager56 commented 1 year ago

Thanks Dr. Stewart and Dr. Shortridge. The porosity at the lowest layer is 52.5%. I agree that this looks off at the lowest layer - PEST was not able to optimize the lowest layer as good as the top layers. This is the layer that we added later on in order to consider the restricting layer present on the field. Dr. Shortridge, I agree that our N leaching values (NO3+NH4) were high then as well for the 0N treatment plot. When I check it right now, the N leaching (NO3+NH4) for the 0N and 180 N treatment are (20+13kg/ha) and (58+17kg/ha), respectively, in 2021 (no irrigation). The N uptake values are 66 and 185 kg/ha, respectively, for 0N and 180N treatments.

I will look into the 2021 simulation results right now to see what might be the issue.

sumanager56 commented 1 year ago

Hi all, 1) I looked into the reasons behind higher N leaching values. One significant factor was the fertilizer ammendment option in SWAP. I was using the Mineral N fertilizer option in SWAP which assumes that NH4-N and NO3-N ratio are 50-50%. However, I tried to see how it will affect the results when using the UAN fertilizer which is actually applied in the fields (UAN is 50% Urea- do not contain NO3-N; 25%NH4-N; and 25% NO3-N). So, I selected two different ammendment options in SWAP (90 kg/ha of Urea with just 0.46 NH4-Nfraction and 90 kg/ha of Mineral fertilizer with 0.5 NH4-Nfraction and 0.5 NO3-Nfraction). Now the proportion of NO3 applied to the field is actually reduced to ~25% instead of ~50% that I was using before (45kg/ha NO3-N and 86.4kg/ha NH4-N). This reduced the simulated yield values as well for different treatments.

2) I went back to see the latest PEST optimization process and re-ran it using the saturated water content values of bottom soil layer in the range from 40-45% and it did optimize well with value of 42% which should address Dr. Stewart's doubts on high field capacity value of 52% (now it is somewhere close to 40%). I also looked into the observed vs simulated soil moisture plot and they are similar to the previous results.

3) For the 2021 observed vs simulated yield values, I ran them for 0N and 180N treatments and here's the result:

2021 simulated yield and leaching values

Non-irrigated | Yield | Nuptake | NO3_leach | NH4_leach | Drymatter_lb/acre -- | -- | -- | -- | -- | -- 0N | 47 | 66 | 18 | 13 | 8625 180N | 187 | 167 | 32 | 17 | 20083

4) Climate and irrigation management Scenario results - the base scenario values are now all less compared to my previous results table <html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

Parameters | Treatments | Scenario_0 | Scenario_1 | Scenario_2 | Scenario_3 | Scenario_4 | Scenario_5 -- | -- | -- | -- | -- | -- | -- | -- Corn_Yield (bu/acre) | Rainfed | 133 | 132 | 133 | 128 | 131 | 138 | Calendar | 107 | 168 | 157 | 171 | 167 | 157 | Precision | 166 | 186 | 181 | 188 | 185 | 181 N-Yield (kg/ha) | Rainfed | 138 | 134 | 124 | 133 | 134 | 144 | Calendar | 124 | 138 | 129 | 142 | 139 | 130 | Precision | 152 | 167 | 161 | 169 | 167 | 161 NO3-Leach (kg/ha) | Rainfed | 33 | 63 | 96 | 66 | 65 | 58 | Calendar | 36 | 45 | 46 | 45 | 46 | 47 | Precision | 21 | 40 | 39 | 41 | 42 | 40

5) N-balance for scenario 0 (rainfed)- example <html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1728 | Volatilisation | 259 | Ammendments | 900 | | Crop Residues | 0 | Plant uptake | 1612 | Crop Residues | 0 | Plant uptake | 1147 Mineralisation | 1432 | Nitrification | 975 | Nitrification | 975 | Denitrification | 90 Influx top | 27 | Outflux | 337 | Influx top | 27 | Outflux | 662 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3187 | Total | 3183 | Total | 1902 | Total | 1899 | | Diff.Stor | 4 | | | Diff.Stor | 3

Scenario-0

Rplot

Scenario-1

Rplot

Scenario-2

Rplot

Scenario-3

Rplot

Scenario-4

Rplot

Scenario-5

Rplot

ryestewart commented 1 year ago

Hi Suman,

Those leaching numbers look more realistic to me, and the changes you proposed seem reasonable.

Is there a CEC option in the model? How did you parameterize it? I think it's assumed that NH4 doesn't leach much, but I remember seeing at least one study looking at NH4 leaching in the coastal plain (if you can't find any studies like that let me know and I'll try to find it).

thanks, Ryan

Ryan D. Stewart, Ph.D. Associate Professor School of Plant and Environmental Sciences Virginia Tech, 241 Smyth Hall, Blacksburg, VA 24061 Phone: (540) 231-0253; Email: @. @.> Zoom: https://virginiatech.zoom.us/my/ryanstewart

On Tue, Sep 5, 2023 at 1:37 PM Suman Budhathoki @.***> wrote:

Hi all,

1.

I looked into the reasons behind higher N leaching values. One significant factor was the fertilizer ammendment option in SWAP. I was using the Mineral N fertilizer option in SWAP which assumes that NH4-N and NO3-N ratio are 50-50%. However, I tried to see how it will affect the results when using the UAN fertilizer which is actually applied in the fields (UAN is 50% Urea- do not contain NO3-N; 25%NH4-N; and 25% NO3-N). So, I selected two different ammendment options in SWAP (90 kg/ha of Urea with just 0.46 NH4-Nfraction and 90 kg/ha of Mineral fertilizer with 0.5 NH4-Nfraction and 0.5 NO3-Nfraction). Now the proportion of NO3 applied to the field is actually reduced to ~25% instead of ~50% that I was using before (45kg/ha NO3-N and 86.4kg/ha NH4-N). This reduced the simulated yield values as well for different treatments. 2.

I went back to see the latest PEST optimization process and re-ran it using the saturated water content values of bottom soil layer in the range from 40-45% and it did optimize well with value of 42% which should address Dr. Stewart's doubts on high field capacity value of 52% (now it is somewhere close to 40%). I also looked into the observed vs simulated soil moisture plot and they are similar to the previous results. 3.

For the 2021 observed vs simulated yield values, I ran them for 0N and 180N treatments and here's the result:

2021 simulated yield and leaching values Non-irrigated Yield Nuptake NO3_leach NH4_leach Drymatter_kg/ha 0N 47 66 18 13 9668 180N 187 167 32 17 22511

Climate and irrigation management Scenario results - the base scenario values are now all less compared to my previous results table

Parameters Treatments Scenario_0 Scenario_1 Scenario_2 Scenario_3 Scenario_4 Scenario_5 Corn_Yield (bu/acre) Rainfed 133 132 133 128 131 138 Calendar 107 168 157 171 167 157 Precision 166 186 181 188 185 181 N-Yield (kg/ha) Rainfed 138 134 124 133 134 144 Calendar 124 138 129 142 139 130 Precision 152 167 161 169 167 161 N-Leach (kg/ha) Rainfed 33 63 96 66 65 58 Calendar 36 45 46 45 46 47 Precision 21 40 39 41 42 40

N-balance for scenario 0 (rainfed)- example

NH4-N NO3-N Input Output Input Output Ammendments 1728 Volatilisation 259 Ammendments 900 Crop Residues 0 Plant uptake 1612 Crop Residues 0 Plant uptake 1147 Mineralisation 1432 Nitrification 975 Nitrification 975 Denitrification 90 Influx top 27 Outflux 337 Influx top 27 Outflux 662 Influx lat 0 Influx lat 0 Influx bot 0 Influx bot 0 Total 3187 Total 3183 Total 1902 Total 1899 Diff.Stor 4 Diff.Stor 3 Scenario-0

[image: Rplot] https://user-images.githubusercontent.com/99036349/265772861-b0e9bac1-73a8-44d3-a6e5-0a1e5bf2abfe.png Scenario-1

[image: Rplot] https://user-images.githubusercontent.com/99036349/265772890-1a789ccd-52f5-4cbd-88f0-3929b4a27fe5.png Scenario-2

[image: Rplot] https://user-images.githubusercontent.com/99036349/265772925-3c3859f7-4b77-4ee1-8a38-1d4631b93f2e.png Scenario-3

[image: Rplot] https://user-images.githubusercontent.com/99036349/265772944-0573bf8a-7496-4ebc-96b0-fb86e2504f45.png Scenario-4

[image: Rplot] https://user-images.githubusercontent.com/99036349/265772966-ff0ee5fa-6351-46e4-9226-23a08f86a9f1.png Scenario-5

[image: Rplot] https://user-images.githubusercontent.com/99036349/265772994-a0384f1d-0fa6-4fa2-aa4b-d5354b9da5f7.png

— Reply to this email directly, view it on GitHub https://github.com/sumanager56/SWAP-WOFOST/issues/15#issuecomment-1707033656, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3TXL5TT7NSLNL6KDUP622TXY5PNRANCNFSM6AAAAAAYMJXBYU . You are receiving this because you were mentioned.Message ID: @.***>

sumanager56 commented 1 year ago

Thanks Dr. Stewart. I could not find a paper that relates to NH4 leaching in the coastal plains - If you could help me find one I would really appreciate that. There's not a CEC option in the model but the closest parameter is the Sorption coefficient of Ammonium.

SWAP model input file information

! SorpCoef : Ammonium sorption coefficient. Ammonium sorption is described by a linear relation. ! The coefficient is expressed in m3 soil water per kg soil. The retardation of the ! ammonium-N migration in soil is expressed as ! Retardation = 1 + sorption coefficient dry bulkdensity / moisture volume fraction ! Values of SorpCoef = 0.0005, dry bulk density = 1200 kg per m3 soil and moisture volume fraction = 0.3 ! leads to Retardation = 3. This value is somewhat high, but within the range of expected values. ! If the retardation of ammonium migration is soils is known, then the coefficient can be ! calculated from: ! Sorption coefficient = (Retardation - 1) moisture volume / dry bulkdensity ! Simulation results appear to have a minor sensitivity to this coefficient.

The default value was 0.0005 m3/kg and I didnt change it before (As far as I remember it didnt change my results). However, it seems to have a significant impact on the results when I change it from 0.0005 to 0.001. The NH4 uptake increased significantly while the leaching reduced significantly. Here's the N balance for a calendar based irrigation treatment under historical climate (Scenario-0) with Sorpcoef=0.0005 and SorpCoeff=0.001.

SorpCoefficient=0.0005

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1728 | Volatilisation | 259 | Ammendments | 900 | | Crop Residues | 0 | Plant uptake | 1454 | Crop Residues | 0 | Plant uptake | 1029 Mineralisation | 1288 | Nitrification | 961 | Nitrification | 961 | Denitrification | 143 Influx top | 32 | Outflux | 375 | Influx top | 32 | Outflux | 720 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3048 | Total | 3049 | Total | 1893 | Total | 1892 | | Diff.Stor | -1 | | | Diff.Stor | 1

SorpCeofficient=0.001

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1728 | Volatilisation | 259 | Ammendments | 900 | | Crop Residues | 0 | Plant uptake | 1832 | Crop Residues | 0 | Plant uptake | 879 Mineralisation | 1288 | Nitrification | 692 | Nitrification | 692 | Denitrification | 125 Influx top | 32 | Outflux | 276 | Influx top | 32 | Outflux | 624 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3048 | Total | 3059 | Total | 1624 | Total | 1628 | | Diff.Stor | -11 | | | Diff.Stor | -4

sumanager56 commented 1 year ago

In response to our discussion on significantly low calendar based corn yield on Scenario-0, here's the N-balance for Calendar and Precision Irrigation treatments. The NH4-N uptake is significantly low for the calendar-based irrigation.

Calendar-Scenario-0

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1728 | Volatilisation | 259 | Ammendments | 900 | | Crop Residues | 0 | Plant uptake | 1454 | Crop Residues | 0 | Plant uptake | 1029 Mineralisation | 1288 | Nitrification | 961 | Nitrification | 961 | Denitrification | 143 Influx top | 32 | Outflux | 375 | Influx top | 32 | Outflux | 720 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3048 | Total | 3049 | Total | 1893 | Total | 1892 | | Diff.Stor | -1 | | | Diff.Stor | 1

Precision-Scenario-0

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 1728 | Volatilisation | 259 | Ammendments | 900 | | Crop Residues | 0 | Plant uptake | 1814 | Crop Residues | 0 | Plant uptake | 1232 Mineralisation | 1348 | Nitrification | 779 | Nitrification | 779 | Denitrification | 57 Influx top | 30 | Outflux | 257 | Influx top | 30 | Outflux | 420 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 3106 | Total | 3109 | Total | 1709 | Total | 1709 | | Diff.Stor | -3 | | | Diff.Stor | 0

ryestewart commented 1 year ago

Hi Suman,

Here's the source about NH4 leaching I was thinking about. It is a student thesis, but there may be some relevant information in the literature review as well.

https://getd.libs.uga.edu/pdfs/barlow_holly_p_200605_ms.pdf

Ryan D. Stewart, Ph.D. Associate Professor School of Plant and Environmental Sciences Virginia Tech, 241 Smyth Hall, Blacksburg, VA 24061 Phone: (540) 231-0253; Email: @. @.> Zoom: https://virginiatech.zoom.us/my/ryanstewart

On Tue, Sep 5, 2023 at 4:37 PM Suman Budhathoki @.***> wrote:

Thanks Dr. Stewart. I could not find a paper that relates to NH3 leaching in the coastal plains - If you could help me find one I would really appreciate that. There's not a CEC option in the model but the closest parameter is the Sorption coefficient of Ammonium. SWAP model input file information

! SorpCoef : Ammonium sorption coefficient. Ammonium sorption is described by a linear relation. ! The coefficient is expressed in m3 soil water per kg soil. The retardation of the ! ammonium-N migration in soil is expressed as ! Retardation = 1 + sorption coefficient dry bulkdensity / moisture volume fraction ! Values of SorpCoef = 0.0005, dry bulk density = 1200 kg per m3 soil and moisture volume fraction = 0.3 ! leads to Retardation = 3. This value is somewhat high, but within the range of expected values. ! If the retardation of ammonium migration is soils is known, then the coefficient can be ! calculated from: ! Sorption coefficient = (Retardation - 1) moisture volume / dry bulkdensity ! Simulation results appear to have a minor sensitivity to this coefficient.

The default value was 0.0005 m3/kg and I didnt change it before (As far as I remember it didnt change my results). However, it seems to have a significant impact on the results when I change it from 0.0005 to 0.001. The NH4 uptake increased significantly while the leaching reduced significantly. Here's the N balance for a calendar based irrigation treatment under historical climate (Scenario-0) with Sorpcoef=0.0005 and SorpCoeff=0.001. SorpCoefficient=0.0005 NH4-N NO3-N Input Output Input Output Ammendments 1728 Volatilisation 259 Ammendments 900 Crop Residues 0 Plant uptake 1454 Crop Residues 0 Plant uptake 1029 Mineralisation 1288 Nitrification 961 Nitrification 961 Denitrification 143 Influx top 32 Outflux 375 Influx top 32 Outflux 720 Influx lat 0 Influx lat 0 Influx bot 0 Influx bot 0 Total 3048 Total 3049 Total 1893 Total 1892 Diff.Stor -1 Diff.Stor 1 SorpCeofficient=0.001 NH4-N NO3-N Input Output Input Output Ammendments 1728 Volatilisation 259 Ammendments 900 Crop Residues 0 Plant uptake 1832 Crop Residues 0 Plant uptake 879 Mineralisation 1288 Nitrification 692 Nitrification 692 Denitrification 125 Influx top 32 Outflux 276 Influx top 32 Outflux 624 Influx lat 0 Influx lat 0 Influx bot 0 Influx bot 0 Total 3048 Total 3059 Total 1624 Total 1628 Diff.Stor -11 Diff.Stor -4

— Reply to this email directly, view it on GitHub https://github.com/sumanager56/SWAP-WOFOST/issues/15#issuecomment-1707277356, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3TXL5UMF5RTRTUBHQ5F2P3XY6ERRANCNFSM6AAAAAAYMJXBYU . You are receiving this because you were mentioned.Message ID: @.***>

julieshortridge commented 1 year ago

Thanks Suman. Regarding your last comment where you showed the N balance between the calendar and precision treatments, are those numbers with a sorption parameter value of 0.0005 or 0.001? If we're thinking that the ammonium leaching values are too high, I could see that being exacerbated under the calendar treatment.

sumanager56 commented 1 year ago

Thank you Dr. Stewart. It's helpful!

Thanks Dr. Shortridge. Those numbers are based on a sorption value of 0.0005. Yes, calendar treatment values are pretty high compared to the precision treatments.

julieshortridge commented 1 year ago

So what happens when you switch them to 0.001?

sumanager56 commented 1 year ago

The yield value goes up for all treatments and both NH4-N and NO3-N leaching goes down, however, the difference between calendar and precision is still huge.

sumanager56 commented 1 year ago

The difference between calendar and precision treatment yields in the Scenario-0 seem to be largely controlled by the N management. When I used the same N application method (split application) for the calendar treatment (instead of one time application at planting), the yield difference dropped down significantly (160 bu/ac for precision vs 140 bu/ac for calendar).

julieshortridge commented 1 year ago

Ah ok, well split application is pretty common anyways so maybe we should just use that for the calendar treatment as well. What I would like to see is, for 2021: 1) observed yield and N uptake for the rainfed, calendar, and precision irrigated trials 2) simulated yield, N uptake, and N leaching for the rainfed, calendar, and precision irrigated treatments (these should the use actual N application dates) 3) N balance tables (2021) for the three treatments.

Once we can confirm that things look reasonable in the 2021 comparison, then we can check the long-term simulations.

sumanager56 commented 1 year ago

Thanks Dr. Shortridge. Here are the results:

Observed yield and N uptake

Ammendments | Irrigation | DRY WEIGHT | N uptake | YIELD -- | -- | -- | -- | -- | | lb/ac | lb/ac | bu/ac 180 N kg/ha | No Irrigation | 23551.1 | 431 | 202.5 180 N kg/ha | No Irrigation | 19904.3 | 175 | 131.4 180 N kg/ha | No Irrigation | 19998.4 | 210 | 190.8 180 N kg/ha | No Irrigation | 17597.4 | 151 | 202.4 Mean = | | 20262.8 | 242 | 181.7 180 N kg/ha | Prec. Irrigation | 23544.6 | 370 | 183 180 N kg/ha | Prec. Irrigation | 23433.9 | 321 | 210.9 180 N kg/ha | Prec. Irrigation | 20084.5 | 171 | 207 180 N kg/ha | Prec. Irrigation | 19221.2 | 196 | 217.1 Mean = | | 21571 | 264 | 204.5 180 N kg/ha | Conv. Irrigation | 19035.6 | 291 | 223.4 180 N kg/ha | Conv. Irrigation | 21642.5 | 266 | 114.2 180 N kg/ha | Conv. Irrigation | 27718.9 | 252 | 225.1 180 N kg/ha | Conv. Irrigation | 20707.7 | 201 | 216.3 Mean = | | 22276.2 | 253 | 194.7

Simulated yield, N uptake and N leaching

Here, calendar irrigation is the same used in the Scenario-0 irrigation. Also, precision irrigation is same as our multi year simulations (50% to 85% of FC).

Simulated | Ammendments | Application | Yield (bu/ac) | Nuptake (kg/ha) | NO3_leach (kg/ha) | NH4_leach (kg/ha) | Drymatter (lb/acre) -- | -- | -- | -- | -- | -- | -- | -- Rainfed | 180N kg/ha | April20 | 194 | 150 | 36 | 10 | 19652 Calendar | 180N kg/ha | April20&June2 | 187 | 145 | 31 | 10 | 19256 Precision | 180N kg/ha | April20&June2 | 199 | 156 | 30 | 8 | 20134 | | | | | | | Rainfed | 0N kg/ha | None | 40 | 47 | 18 | 7 | 7153 Calendar | 0N kg/ha | None | 30 | 37 | 19 | 7 | 5696 Precision | 0N kg/ha | None | 34 | 45 | 18 | 7 | 6889

N-balance [Rainfed]

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 86.4 | Volatilisation | 0 | Ammendments | 45 | | Crop Residues | 0 | Plant uptake | 86 | Crop Residues | 0 | Plant uptake | 64 Mineralisation | 81 | Nitrification | 62 | Nitrification | 62 | Denitrification | 2.86 Influx top | 1.45 | Outflux | 10 | Influx top | 1.45 | Outflux | 36 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 168.85 | Total | 158 | Total | 108.45 | Total | 102.86 | | Diff.Stor | 10.85 | | | Diff.Stor | 5.59

N-balance [Calendar]

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 86.4 | Volatilisation | 6.39 | Ammendments | 45 | | Crop Residues | 0 | Plant uptake | 87 | Crop Residues | Plant uptake | 58 Mineralisation | 73 | Nitrification | 53 | Nitrification | 53 | Denitrification | 6 Influx top | 1.63 | Outflux | 10 | Influx top | 1.63 | Outflux | 31 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 161.03 | Total | 156.39 | Total | 99.63 | Total | 95 | | Diff.Stor | 4.64 | | | Diff.Stor | 4.63

N-balance [Precision]

NH4-N | | | | NO3-N | | | -- | -- | -- | -- | -- | -- | -- | -- Input | | Output | | Input | | Output | Ammendments | 86.4 | Volatilisation | 6.39 | Ammendments | 45 | | Crop Residues | 0 | Plant uptake | 96 | Crop Residues | Plant uptake | 59 Mineralisation | 80 | Nitrification | 51 | Nitrification | 51 | Denitrification | 3 Influx top | 1.5 | Outflux | 8 | Influx top | 1.5 | Outflux | 30 Influx lat | 0 | | | Influx lat | 0 | | Influx bot | 0 | | | Influx bot | 0 | | Total | 167.9 | Total | 161.39 | Total | 97.5 | Total | 92 | | Diff.Stor | 6.51 | | | Diff.Stor | 5.5

sumanager56 / SWAP-WOFOST