Calibration and validation of ceres rice crop simulation model
| dc.contributor.advisor | Girija Devi, L | |
| dc.contributor.author | Anju, V S | |
| dc.contributor.author | KAU | |
| dc.date.accessioned | 2020-02-29T10:19:10Z | |
| dc.date.available | 2020-02-29T10:19:10Z | |
| dc.date.issued | 2018 | |
| dc.description | Doctor of Philosophy in Agriculture | en_US |
| dc.description.abstract | The project entitled ‘Calibration and validation of CERES-Rice crop simulation model’ was conducted in the Department of Agronomy, College of Agriculture, Vellayani from 2016 to 2018 with the objectives to calibrate and validate CERES-Rice model to generate the genetic coefficients of the rice variety Prathyasa, to study the crop- weather relationship and to quantify the yield gap of the variety by running simulations. The field experiment was conducted at Upanniyoor Panchayat in farmer’s field for four seasons (Virippu 2016 and 2017 and Mundakan 2016 and 2017) and it was laid out in randomized block design. The treatments consisted of five dates of sowing each in Virippu (D1 - May 31, D2 –June 15, D3 –June 30, D4 –July 15 and D5 – July 30) and five dates of sowing each in Mundakan (D1 -Oct 14 / Sept 26, D2 – Oct 30 /Oct 10, D3 – Nov 14/ Oct 25, D4 –Nov 30/Nov 9 and D5 – Dec 14/Nov 24). The sowing dates in Mundakan seasons of 2016 and 2017 varied due to the delayed onset of rainfall in 2016. The plot size was 5 x 4 m2 with three replications. Routine observations on height, leaf area, dry matter production (DMP), number of tillers, panicles, spikelets per panicle, filled grains per panicle, 1000 grain weight, straw yield and grain yield were recorded apart from phenological observations. Soil analysis was conducted before and after the experiment. The soil and crop data collected from the experimental field and weather data from the Department of Agrometeorolgy were used as inputs for running the model. Study on phenology revealed that the crop duration decreased from 111 to 100 and 117 to 107 days respectively in Virippu 2016 and 2017. A similar decreasing trend was observed in Mundakan 2016, but in Mundakan 2017, it increased from 114 to 117 days in early sowing and decreased drastically from 117 to 105 days in delayed sowing. The height of the plant was found varying at different stages, D1 produced the tallest plants at harvest in Virippu seasons of both the years, while it was the highest in D3 in Mundakan 2016 at different stages and D5 in Mundakan 2017. The number of tillers was the highest in D2 and D1 respectively in Virippu in both the years and D2 and D1 respectively in Mundakan 2016 and 2017. The DMP was the highest in D2 and D1 respectively in Virippu and Mundakan 2016 and 2017. The grain yield was the highest in D2 in both the seasons in 2016 and D1 in both the seasons in 2017. The yield attributes such as productive tillers m-2 was the highest in D2 in both the seasons in 2016 and D1 in both seasons in 2017. The number of spikelets per panicle was the highest in D1 during Virippu 2016 and 2017 and Mundakan 2017 and D2 in Mundakan 2016. D1 in Virippu 2017 was on par with D2, and D2 in Mundakan 2016 was on par with D1. The number of filled grains per panicle was higher in D1 in Virippu 2016 and Mundakan 2017, while D2 recorded higher filled grains per panicle in Virippu 2017 and Mundakan 2016. D1 in Virippu 2016 was on par with D2 and D2 in Mundakan 2016 was on par with D1. The harvest index (HI) was higher in D1 in Virippu 2016 and 2017 and Mundakan 2017, while D2 recorded higher HI in Mundakan 2016. D1 in Virippu 2016 was on par with D2 and D5, and D1 in Virippu 2017 was on par with D2. In Mundakan season, D2 was on par with D1 and D5 in 2016 and D1 was on par with D2 in 2017. In Virippu and Mundakan 2016, N uptake was the highest in D1 while P and K uptake were the highest in D2, whereas in Virippu and Mundakan 2017, N, P and K uptake were the highest in D1. The organic carbon content of the soil was found influenced only after Virippu 2016 with D5 recording the highest value. In the case of available N, P and K status of the soil, only the N status was found affected and that was only after Mundakan 2017 with D2 recording the highest value. Crop weather relationship was studied by computing the different heat units such as Growing degree days (GDD), Heliothermal units (HTU), Photothermal units (PTU) and Heat unit efficiency (HUE) at different stages such as sowing to active tillering (P1), active tillering to panicle initiation (P2), panicle initiation to booting (P3), booting to heading (P4), heading to 50% flowering (P5), 50% flowering to physiological maturity (P6), vegetative stage (P7), reproductive stage (P8) and ripening stage (P9). These heat units computed were the highest in D1 and showed positive correlation with yield for GDD at P1, HTU at P5 and P6, PTU at P1 in Virippu, while in Mundakan positive correlation was obtained with GDD at P1 and P7, HTU at P2, PTU at P1 and P7. Negative correlation was obtained with GDD at P3, P8 and P9 and PTU at P4, P8 and P9 in Virippu and with HTU at P2 and P3 and PTU at P6 and P9 in Mundakan. The correlation between yield and yield attributes with weather parameters revealed positive correlation for minimum temperature at P3, P4 and P8, RH I & RH II at P1 and P7, BSS at P2, P3, P4 and P6, rainfall and rainy days at P1 and P7, pan evaporation at P6, P8 and P9 and wind speed at P6 and P9 in Virippu season. Negative correlation was observed with minimum temperature, pan evaporation and wind velocity at P1, rainfall at P3, P4, P6, P8 and P9, rainy days at P6, P8 and P9, RH I at P6, P8 and P9, RHII at P5, P6 and P9 in Virippu season. In Mundakan season positive correlation was obtained with maximum temperature from P1 to P9 except P6, RH I at P2, P4, P5, P7, P8 and P9 and rainy days at P1 and negative correlation with maximum temperature at P1, P2 and P6, minimum temperature at P6, BSS at P2, rainfall at P6, pan evaporation at P3, P5, P6, P7, P8 and P9. The genetic coefficients for the variety Prathyasa was generated by calibrating the CERES-Rice model by using the data of Virippu rice 2016 and validated by using the data of Mundakan 2016, Virippu and Mundakan 2017 respectively and the genetic coefficients generated were P1-720, P2R-33.7, P5-21.3, P2O-12, G1-38.7, G2-0.028, G3-1, G4-1 respectively. Model simulated results showed that there was close association between observed and simulated yield and the error percentage varied from -16.90 to 16.55 for Virippu 2016 and from -1.26 to 64.77 in Virippu 2017. In Mundakan, error per cent ranged from -13.43 to 16.63 in 2016 and from -11.09 to 12.58 in 2017. The error percentage for panicle initiation day varied from 1.96 to 18.37 in Virippu 2016, while it varied from -5.88 to 10 in Virippu 2017 and for Mundakan it varied from -8.16 to 0 in 2016 and from -8 to 1.96 in 2017. Similarly the error percentage for anthesis day varied from 8.54 to 11.25 and 4.88 to 9.88 in Virippu 2016 and 2017 and from 4.88 to 9.88 and 3.66 to 8.43 in 2016 and Mundakan 2017. The error percentage of physiological maturity day varied from -1.96 to 18 in Virippu 2016, while it deviated from -7.84 to 0.99 in Virippu 2017. During Mundakan season, error percentage ranged from -1.98 to 6.54 in 2016 and from -2.91 to 4.81 in 2017. Regression equations for grain yield were developed for certain phenological stages in Virippu and Mundakan from highly correlated weather parameters. The yield gap quantification revealed that the highest total and sowing yield gaps were in delayed sowing (D5), management yield gap in early sowing(D1), and the lowest in D3 and D4 (delayed sowing) and D2 (early sowing) respectively for the same parameters. Thus, the study enabled to generate the genetic coefficients of variety ‘Prathyasa’ and simulated the grain yield and panicle initiation, anthesis and physiological maturity days with minimum error percentage. The study also helped to quantify various yield gaps such as total, management and sowing gaps due to different dates of sowing, from the potential yield generated by the model along with the attainable and actual yield data supplied from the field experiment and farmers’ field. The various correlations worked out between yield, weather parameters and heat units provided an insight into the crop weather relationship. Finally, and the foremost implication of the study is that delayed sowing reduces the yield considerably in rice crop in both the seasons irrespective of other factors. | en_US |
| dc.identifier.uri | http://krishikosh.egranth.ac.in/handle/1/5810144007 | |
| dc.keywords | Agriculture, Manures, Tillering, Wind Speed, Humidity, Harvest | en_US |
| dc.language.iso | en | en_US |
| dc.pages | 211 | en_US |
| dc.publisher | Department of Agronomy College of Agriculture, Vellayani | en_US |
| dc.sub | Agriculture | en_US |
| dc.subject | null | en_US |
| dc.theme | to calibrate and validate CERES-Rice model to generate the genetic coefficients of the rice variety Prathyasa, to study the crop- weather relationship and to quantify the yield gap of the variety by running simulations. | en_US |
| dc.these.type | Ph.D | en_US |
| dc.title | Calibration and validation of ceres rice crop simulation model | en_US |
| dc.type | Thesis | en_US |