Abstract: Maize (Zea mays L.) is the most prominent crop in the Northeast Farming Region of China (NFR), and drought has been the largest limitation for maize production in this area during recent decades. The question of how to adapt maize production to drought has received great attention from policy makers, researchers and farmers. In order to evaluate the effects of adaptation strategies against drought and examine the influences of policy supports and farmer households’ characteristics on adopting decisions, a large scale household survey was conducted in five representative maize production counties across NFR. Our survey results indicated that using variety diversification, drought resistant varieties and dibbling irrigation are the three major adaptation strategies against drought in spring, and farmers also adopted changes in sowing time, conservation tillage and mulching to cope with drought in spring. About 20% and 18% of households enhanced irrigation against drought in summer and autumn, respectively. Deep loosening tillage and organic fertilizer are also options for farmers to resist drought in summer. Maize yield was highly dependent on soil qualities, with yields on land of high soil quality approximately 1050 kg/ha and 2400 kg/ha higher than for normal and poor soil conditions, respectively. Using variety diversification and drought resistant varieties can respectively increase maize yield by approximately 150 and 220 kg/ha under drought. Conservation tillage increased maize yield by 438–459 kg/ha in drought years. Irrigation improved maize yield by 419–435 kg/ha and 444–463 kg/ha against drought in summer and autumn, respectively. Offering information service, financial and technical support can greatly increase the use of adaptation strategies for farmers to cope with drought. However, only 46% of households received information service, 43% of households received financial support, and 26% of households received technical support against drought from the local government. The maize acreage and the irrigation access are the major factors that influenced farmers’ decisions to apply adaptation strategies to cope with drought in each season, but only 25% of households have access to irrigation. This indicates the need for enhanced public support for farmers to better cope with drought in maize production, particularly through improving access to irrigation.

Abstract: The aim of this paper is to provide a critical analysis of the strengths and weaknesses of a non-standardised Model Requirements Analysis (MRA) used for the purpose of developing the Sustainability Impact Assessment Tool (SIAT). By ‘non-standardised’ we mean not strictly following a published MRA method. The underlying question we are interested in addressing is how non-standardised methods, often employed in research driven projects, compare to defined methods with more standardised structure, with regards their ability to capture model requirements effectively, and with regards their overall usability. Through describing and critically assessing the specific features of the non-standardised MRA employed, the ambition of this paper is to provide insights useful for impact assessment tool (IAT) development. Specifically, the paper will (i) characterise kinds of user requirements relevant to the functionality and design of IATs; (ii) highlight the strengths and weaknesses of non-standardised MRA for user requirements capture, analysis and reflection in the context of IAT; (iii) critically reflect on the process and outcomes of having used a non-standardised MRA in comparison with other more standardised approaches. To accomplish these aims, we first review methods available for IAT development before describing the SIAT development process, including the MRA employed. Major strengths and weaknesses of the MRA method are then discussed in terms of user identification and characterisation, organisational characterisation and embedding, and ability to capture design options for ensuring usability and usefulness. A detailed assessment on the structural differences of MRA with two advanced approaches (Integrated DSS design and goal directed design) and their role in performance of the MRA tool is used to critique the approach employed. The results show that MRA is able to bring thematic integration, establish system performance and technical thresholds as well as detailing quality and transparency guidelines. Nevertheless the discussion points out to a number of deficiencies in application – (i) a need to more effectively characterise potential users, and; (ii) a need to better foster communication among the distinguished roles in the development process. If addressed these deficiencies, SIAT non-standardised MRA could have brought out better outcomes in terms of tool usability and usefulness, and improved embedding of the tool into conditions of targeted end-users. (C) 2013 Elsevier Ltd. All rights reserved.

Abstract: This paper deals with the problem of predicting biomass and grain protein content using improved particle filtering (IPF) based on minimizing the Kullback-Leibler divergence. The performances of IPF are compared with those of the conventional particle filtering (PF) in two comparative studies. In the first one, we apply IPF and PF at a simple dynamic crop model with the aim to predict a single state variable, namely the winter wheat biomass, and to estimate several model parameters. In the second study, the proposed IPF and the PF are applied to a complex crop model (AZODYN) to predict a winter-wheat quality criterion, namely the grain protein content. The results of both comparative studies reveal that the IPF method provides a better estimation accuracy than the PF method. The benefit of the IPF method lies in its ability to provide accuracy related advantages over the PF method since, unlike the PF which depends on the choice of the sampling distribution used to estimate the posterior distribution, the IPF yields an optimum choice of this sampling distribution, which also utilizes the observed data. The performance of the proposed method is evaluated in terms of estimation accuracy, root mean square error, mean absolute error and execution times.