UK_CHINA Joint Project to Assess the Potential Impacts
of Climate Change on Chinese Agriculture
Interim Progress Reports
(Sep. 1, 2001 – Aug. 31, 2002)
Under the Statement on Joint Work on Climate Change Research by MOST of China and DEFRA of UK, based on the Contracts of Agro-meteorology Institute with AEA and AMI staff with Hadley Center, some valuable progresses have been made since the signatures. The capacity of Hadley Center HadAM3H to simulate present climate over China region had been validated with observation and CRU datasets. It becomes easier to modify national climate change scenarios for the 2020s, 2050s and 2080s by RCM and expertise from the Hadley Center. National socio-economic scenarios for the 2020s and 2050s relevant to agriculture have been developed. The regional crop model (ICCC) for whole China, which rooted in CERES Crop model, hasve been formed for further simulation and assessment. A land-use change scheme for China is developing, involving the application of remote sensing data. The framework of an integrated model has been made to Chinese case study. Four Chinese young scientists and one study group of other 4 researchers were in UK for their first round training on regional climatic modeling, crop modeling, GIS use, integrated modeling, soil carbon modeling and climate change sciences and policy. The task of providing an overview of the overall effects of climate change on agriculture in China, including economic costs of damages and/or adaptation, has shown a significant progress.
1. Establishment of database:
The crop modeling working group is collecting data they need for running the models. The following databases are completed or under development:
(1) Climatic database: We established climatic database, which includes daily maximum, minimum, mean temperatures, daily precipitation, daily sunshine, radiation, wind speed, and humidity for 400 stations from the beginning of the station establishment to the end of 1999. The total number of data is about 50 million.
(2) Soil database: The soil database was completed. The organic carbon content, pH, soil moisture, field capacity, bulk density were collected. The total number of data is t about 580 thousand.
(3) Social economic database: We are establishing the social economic database, which include population, GDP, rural population, crop planting area, crop yields, grain price, and so on for 1990 and 1991. The total number was estimated to be 400 thousands. The database for other years is under development.
(4) Crop characteristics database: In order to calibrate CERES crop model and determine genetic coefficients, the agro-meteorological, management and crop yield data, like variety characteristics, planting date, phonological date, management practice, have been collected for three years (1998-2000), at 400 crop stations of 31 province all over China, including wheat, maize, cotton and rice. Also, the data for 20 years (1981-2000) for 31 stations are also collected for genotype evolution analyses, there are 14 stations for maize, 13 stations for rice and 9 stations for cotton. All the stations are selected in main crop areas or sensitive climate change areas. Totally, we have copied nearly 5,000 pages for the crop data. An observation database is setting up. This database is very useful to validate crop models
2. Validation and Climate Change Responses of Hadley Centre GCMs over China Region
The capacity of Hadley Centre GCM to simulate present climate over China region had been validated with observation and CRU datasets. It is indicated from a series of analyses of model's evolution with the datasets of HadCM2, HadCM3, and HadAM3H that Hadley Centre GCMs performs well over East China, and the high-resolution GCM-HadAM3H can improve the simulation greatly, especially over Northwest China and Tibetan Plateau. It is presented from the seasonal cycle of surface air temperature and precipitation that the temperature cycle patterns of HadAM3H are improved obviously from those of HadCM2; on the contrary, the simulated patterns of precipitation are also improved a lot, but not so close to the observed ones. It is likely from the scenario analyses that the precipitation would increase over the catchment of the Yangzi River and the North China Plain, which could relief the drought in the North China Plain, but increase the risk of flooding over the Yangzi River.
3. RothC model
Dr Guo has been in UK since Jan 29 to receive training on soil model in Institute of Arable Crops Research (IACR)-Rothamsted. Dr Guo works with Dr. Pete Falloon and Dr. Kevin Coleman on soil carbon model-RothC model. RothC model was earlier developed by Jenkinson and Rayner, and by Hart, then been revised by Dr. David Jenkinson and Dr. Kevin Coleman. It is used for the turn over of organic carbon of top soil in non-waterlogged soil. This model was originally developed appropriate for arable soils in the temperate zones. It has been validated in different countries, such as England, Germany, USA, Australia, Czech etc. While it can hardly be validated in China.
RothC model uses a monthly time step to calculate total organic carbon, microbial biomass carbon, and radiative 14C on a years-to-centuries timescale. It needs the following data input: monthly rainfall, monthly open pan evaporation, average monthly air temperature, clay content of the top soil, estimate of the decomposability of incoming plant material, soil cover, monthly input of plant residues, monthly input of farmyard manure.
Dr. Guo has mastered the basic technique for running this model through one month’s training in UK. She has gotten the appropriate rate of one of the most important parameters needed by the model---input of plant residue--- which varies in different climate zones and different soil conditions. This is acquired by using actual measured data in China to validate the simulated results. This rate is about 1.1-1.5 t C/ha/y for winter wheat-summer maize rotation system in China. This rotation pattern is the most popular rotation pattern in North and Central China. This rate is somewhat lower compared with 1.3-1.8 t C/ha/y for the winter wheat rotation pattern of Rotahmsted fields. Dr. Guo got this parameter through the validation model work employing a 13 years long-term experiment in China after iteratively estimate to fit the measured data to the simulated data. For the next step, Dr Guo will focus on the regional soil carbon model—to learn the data requirement and sorting of regional model, and validate more site model.
Besides the RothC model, there are other famous models, such as CENTURY model and DNDC model which could simulate the turn over of soil organic carbon. To our understanding, RothC model is somewhat weaker in some aspects on simulating soil carbon. It needs a manual estimated parameter (i.e. plant input), while it does not include a sub-model of plant growth to produce this rate, which is unknown for most of the experiments. It could not produce this parameter automatically through model itself. While, other models such as CENTURY model have a sub-model of plant production. There is no manual estimate on this unknown parameter for CENTURY model. Dr. Guo expects to combine other models to estimate the soil carbon stocks in his future work.
Dr Guo has gotten 1:4,000,000 digital soil type map and almost all the documents published from 1990-2000 on the GHG emission flux in China and is classifying them according to different rotation patterns and different climate zones.
4. Crop modeling:
We have collected some models to simulate the crop yield under the climate change. These crop models include rice, wheat and corn. As to these models, we select CERES series. It has simple structure, a good deal of function and the friendly interface. It can simulate the crop yield both with irrigation or without irrigation, and can simulate the growth stage and organic development of different crop variety under different climate conditions. It consists of four sub-programs: soil-water sub-program, nitrogen balance sub-program, phenology sub-program and crop growth sub-program. Besides, it has some other functions, such as the calendar conversion assist-program, the print output assist-program and so on.
1) Collect the model code and find out the model mechanism. In order to simulate the yield grid by grid all over the country, we must do some betterment on the model. Such as let it can do regionally, automatically. So we collect the model code of version 1.1 to version2.1, hoping during the training in UK we can develop a fine regional model.
2) Modify the model parameter and optimize the model. In order to get the finest simulated results, we use the field experiment to adjust the model parameter. Through that, the model would reflect the local condition more deeply.
3) Develop other database for input into the model, those are weather data, soil data, genetic coefficient, economic character and so on, weather data includes daily maximum air temperature, daily minimum air temperature, daily precipitation and daily solar radiation ideally covering the period 1950-1999 to allow for adequate characterization of the baseline climate, including inter-annual variability and generation of the climate change scenarios. Soil data composed of horizon thickness, water content at saturation, water content at drained upper limit, water content at lower limit of plant extraction, initial soil water content, soil albedo, stage soil evaporation, profile drainage coefficient, runoff curve number, root length per unit root weight, rate of root depth increase, root growth preference factor to allocate roots by horizon, pH, organic matter etc. the genetic coefficient and temperature to tolerance coefficient.
4) Understood the regional crop model’s principles, and fulfilled it on the rice one.
Map1: Structure map of ICCCC model
5. Social Economic modeling
There are 3 socio-economic scenarios under 2020s and 2050s respectively. The mail parameters in Socio-economic scenarios include population scenarios and urbanisation scenarios. The population increase and urbanization have a major effect upon agricultural land area.
How to predict the population density in each grid is very difficult. There are two kind methodologies can be employed to reallocate the population based on the population increase and urbanization scenarios. One is regionalization based on the economic development, living standards, climate and so on. The population density distribution will be described regionally based on population and urbanization scenarios. Another methodology is to reallocate the population in the cities and towns. The population density in towns and cites are and the areas of cites and town will increase.
Incorporating the population and urbanization scenarios in each grid into the land use scheme to predict land use changes and the agricultural land area in each grid. We are collecting the data and invited some researchers from State Development Research Center to provide the data and suggestions for the Social economic scenarios. We identified the model we will use in the future research. Now we are collecting the data such as population, GDP, grain yield,
6. Integrated assessment modeling
Agriculture is a very important sector for China's economy. It supports almost a quarter of the world population successfully. But it is affected by a combination of human and natural factors. Among them, many are posing a threat to agriculture such as land degradation, growing population with rapid urbanisation, and climate change. The first two factors will decrease the agricultural land area. The later, climate change, because of the increase in temperature, changes in precipitation and the frequency in extreme climate events, will have a direct influence both on agricultural area and crop productivity. The increase in temperature will increase the evaporation rate, thus decreasing the moisture in the soil available for crop growth and speeding up the degradation of soil. These will have a indirect effect on agricultural area and crop productivity.
The previous studies were only focused on the impact of climate change (temperature, precipitation and CO2 increases) on crop productivity. These studies were limited to analyse the first order biophysical effects. The implication of socio-economic trends, cross-sectoral integration of impacts, and adaptation were not incorporated in the impact assessment studies. The impact of these factors on agriculture may be much more important than climate change. The conclusions drawn from these studies can not be properly used to make a recommendation to the policy-makers. Therefore, integrated impact assessment of climate change on agricultural is increasingly recognised as an important technique for assessment the impact of climate change on agriculture.
Two points worth emphasising about integrated model of impact assessment of climate change on agriculture. One is comprehensive, which means that all relevant aspects which might affect agriculture should be incorporated. Another is practical, which means it can be used as a tool to conduct the integrated assessment quantitatively.
The integrated modelling framework consists of drivers (climate related and non-climate related), and models (crop models, land use model, water resource model, soil model, pest and disease model and biodiversity model). ARC/GIS is the linkage between the models and the drivers and among the models. The ARC/GIS also provide the platform for the demonstration of the simulated results. Relevant spatial and non-spatial data, including data from current climate, future climate scenarios, socio-economic scenarios, soil data and crop management data are put into and extracted from ARC/GIS; Using data stored in GIS, the water resource model determines the water supply for irrigation, soil model determines the soil properties, pest and disease model determines the pest and disease out break frequency and intensity, land use model determine the agricultural area in each grid square under climate change scenarios. The results from these models together with the climate scenarios are as inputs to crop models. There are also interconnections among the models.
Potential agricultural area in 2020s and 2050s;
Yield change percentage in 2020s and 2050s;
Production in 2020s and 2050s;
Soil conditions in 2020s and 2050s;
Water supply for irrigation;
Adaptive strategies, cost, effectiveness analysis;
Distribution of vulnerable area by crops, overall vulnerable analysis in 2020s and 2050s;
Conclusions: the impact of climate change on Chinese agriculture; Is the food production if enough for feeding the population in 2020s and 2050s under the climate change context? Adaptation strategies and the cost/benefit of the adaptation measures? Recommendation to policy makers.
7. Crop , soil modeling and integrated assessment modeling training
Dr. Xiong Wei and Li Yue were trainined in Crandfield University for the crop and integrated assessment modeling for 3 and 5 months separately. Dr. Guo Liping was trained in Rothamsted for Soil modeling for 3 months. All of them are well trained and made a significant progress for their tesk of the project. Other 4 scientists from AMI visited UK in the end of September 2001 for climate change. They join the IPCC meeting and visit some institutes and universities and improve their knowledges for this issue.
8. First workshop
The first workshop will be arranged at the Sept 11-12, 2002. The aim of the workshop is to report the progress of the project, to find the problems in the implementation, if any, and get guidance from UK side. The participants for the workshop will include officers from DEFRA and from Ministry of Science and Technology China, and the participants of the project team. Some of UK Scientists are invited to attend the workshop. We also will invite some other Chinese scientists related to the climate change impact assessment area to attend the workshop. The guest speakers both from UK and China would present their research results and exchange their experience. We are very honoured that Vice Ministers Li Xueyong of MOST and Dr. Denis MacShane of UK will present the workshop and give speeches
We have purchased 4 computers, 2 portable computers, copying machine, workstation, and scanner. We are preparing to purchase for A0 printer, GIS software, GIS workstation and web server and other necessary facilities for the workshop.
10. Phase 2 - Case studies
To undertake detailed case studies of the impacts of climate change on agriculture, for selected regions (e.g. N. China), a framework has bee formed based on the results and the interactions with other sectors such as biodiversity, water resources and coastal zones.
This work would be expected to involve collaborating with UK scientists, such as those participating within the UK Climate Impacts Programme.
Watching Slides: http://www.ami.ac.cn/Sino_UK/workshop/Lin_02progress2_Firstday.ppt