Journal of Earth Science & Climatic Change
Open Access

Yield forecasts, Climate change, Crop model

Introduction

In the field of crop yield forecasting [1] or as a climate change impact assessment tool [2-3] in such applications lack of long-term daily weather data and especially data on global solar radiation (RG) has often been a significant challenge, since RG is an indispensable input variable for the majority of these models that simulate crop growth, because photosynthesis itself is primarily driven by solar radiation. Moreover, global solar radiation is a key input variable for methods used at present for potential and actual evapotranspiration estimation, which is an essential part of water balance subroutines in almost all crop models. Other types of mathematical models for crop growth and development apart from the process-oriented methods rely frequently on RG as one of the key independent variables [4] and their outputs might be clearly affected by any RG bias. It has been noted many times [5] that continuous records of global solar radiation measurements are spatially scarce and that the ratio between the number of stations observing daily RG and those measuring temperature and precipitation is highly variable from less than 1:10. The spatial density of weather stations providing complete climatic data sets for crop model inputs is therefore inadequate and the missing RG data have to be substituted. Missing daily radiation data at a given site can either be substituted by data measured at a nearby station, estimated by remote sensing techniques Stochastically generated data may be useful for exploring possible model scenarios for an average theoretical situation using long-term simulations, but they cannot be used for model validation and simulation analysis during a particular period of time [6]. It should also be remembered that the weather generator use would require a time series of observed data (several years at least) in order to set statistical parameters for a particular site. Widespread application of the other two mentioned methods, i.e. linear interpolation and use of neural network analysis in crop growth modeling, is limited for the same reasons as in the weather generator approach and in some cases yield worse results than empirical models [7]. Despite dramatic developments in remote sensing techniques that have made vast databases of RG data available on an unprecedented scale to users, empirical models estimating RG from commonly measured meteorological variables remain an important tool in many agro meteorological applications. Numerous formulas have been made available in previous decades, which might make selection of the most appropriate method for a particular purpose and site. Some of these methods have already been incorporated into crop models as an integral part of the software package, like in case of STICS or SWAP models. Others are developed as standalone applications for use in crop model studies in order to facilitate the preparation of the necessary weather data. All of this poses a problem for some crop model applications (e.g. crop growth monitoring) requiring input parameters to be available in near real time. The second part of the study is thus focused on testing the uncertainty in regional crop model outputs assuming two likely scenarios assuming availability of 1daily extreme temperatures and total precipitation data and 2 only daily extreme temperatures [8-9]. The methods based on cloud cover were not considered in this part of the study as these data were not available in our database for a sufficient number of stations and because with the increasing number of automated weather stations this parameter is not measured. The overall objective was to quantify the uncertainty arising from the use of various RG methods and to compare it with the results found at the local level.

Result and Discussion

We found that the results varied greatly with the soil type and that the majority of the tested RG methods produced statistically significant differences compared with the control runs. As expected, performed best of all methods and the crop model outputs in most cases did not significantly differ from the simulations based on the observed RG. We found that the differences in crop model outputs caused by the RG were more pronounced in fluvisol [10]. The low sensitivity of sandy chernozem can be explained by the much higher crop water stress on this particular soil that was noted during most of the 97 simulated seasons. As soon as water stress becomes the dominant limiting factor, it suppresses growth processes and decreases the overall effect of differences in RG values [11-13].

Conclusion

Therefore a detailed site analysis of the error propagation was made first using two crop models, namely CERES-Barley and CERESWheat. We found that even the method yielding the lowest bias in RG estimates influences a number of key crop model outputs. The precision of the yield estimates and other crop model outputs is lower but still acceptable when estimates based on the diurnal temperature range and cloud cover are used. Methods based on the diurnal temperature range and total precipitation yielded better model estimates than those based on temperature extremes, but there was a systematic bias in the spring barley and winter wheat yields and the considerable increase in seasons with yield departure of more than 25 percent.

Acknowledgment

I’m very glad and thankful for my professor James and family for their support throughout my work. I would also thank my organization.

Conflict of Interest

None.

1. Haregeweyn N, Tsunekawa A, Nyssen J (2015) Soil erosion and conservation in Ethiopia: a review. Prog Phys Geogr 39(6):750-774.

Indexed at, Google Scholar, Crossref

2. Mekuria WA (2017) Assessing the effectiveness of land resource management practices on erosion and vegetative cover using GIS and remote sensing techniques in Melaka watershed, Ethiopia. Environ Syst Res 6:16.

Indexed at, Google Scholar, Crossref

3. Mekuria W, Veldkamp E, Haile M (2007) Effectiveness of exclosures to restore degraded soils as a result of overgrazing in Tigray, Ethiopia. J Arid Environ 270-284.

Indexed at, Google Scholar, Crossref

4. Rosenzweig C (2008) Attributing physical and biological impacts to anthropogenic climate change. Nature 353-357.

Indexed at, Google Scholar, Crossref

5. Shaxson F, Barber R (2003) Optimizing soil Moisture for Plant Production. FAO Soils Bulletin 1-125.

Indexed atGoogle Scholar

6. Shiferaw B, Holden S (2001) Farm-level benefits to investments for mitigating land degradation: Empirical evidence from Ethiopia. Environ Dev Econ 335-358.

Indexed at, Google Scholar, Crossref

7. Tefera B (2002) Nature and causes of land degradation in the Oromiya Region: A review. Work Pap 36.

Indexed at, Google Scholar

8. Adimassu Z, Mekonnen K, Yirga C, Kessler A (2014) Effect of soil bunds on run-off, soil and nutrient losses, and crop yield in the central highlands of Ethiopia. Land Degrad Develop 554-564.

Indexed at, Google Scholar, Crossref

9. Dagnew DC, Guzman CD, Zegeye AD (2015) Impact of conservation practices on runoff and soil loss in the sub-humid Ethiopian highlands: the Debre Mawi watershed. J Hydrol Hydromech 210-219.

Indexed at, Google Scholar, Crossref

10. Fu B, Wang J, Chen L (2003) The effects of land use on soil moisture variation in the Danangou catchment of the Loess Plateau, China. Catena 197-213.

Indexed at, Google Scholar, Crossref

11. Muchena FN (2008) Indicators for Sustainable Land Management in Kenya’s Context. GEF Land Degradation Focal Area Indicators, ETC-East Africa. Nairobi Kenya.

Google Scholar

12. Niyogi D, Kishtawal C, Tripathi S (2010) Observational evidence that agricultural intensification and land use change may be reducing the Indian summer monsoon rainfall. Water Resour Res 46.

Indexed at, Google Scholar, Crossref

13. Nyssen J, Clymans W, Descheemaeker K, Poesen J, et al. (2010) Impact of soil and water conservation measures on catchment hydrological response—a case in North Ethiopia. Hydrol Process 1880-1895.

Indexed at, Google Scholar, Crossref