Alice G. Laborte

Integration of Systems Network (SysNet) tools for regional land use scenario analysis in Asia

Abstract This paper introduces the approach of the Systems research Network (SysNet) for land use planning in tropical Asia with a focus on its main scientific–technical output: the development of the land use planning and analysis system (LUPAS) and its component models. These include crop simulation models, expert systems, GIS, and multiple goal linear programming (MGLP) models for land evaluation and optimization. LUPAS was designed as a decision support system (DSS) for strategic land use planning. Integration of LUPAS components in four case studies was performed in a network with national research teams and local stakeholders. This network allowed iterative evaluation and refinement of LUPAS for scenario analysis on technical and policy changes. Several interactive sessions with stakeholders led to more detail in scenarios (goals and constraints), model features and databases. To facilitate negotiation among stakeholders, the MGLP user interface (UI) was developed. In interactive sessions, goal restrictions are tightened to quantify trade-offs between conflicting goals. Choice and degree of tightening reflect the specific priorities for sustainable land use. The development of LUPAS is exemplified for one case study, Ilocos Norte, Philippines. Weak points of the system include inadequate spatial differentiation of socio-economic characteristics, scarce database for quantifying perennials and mixed cropping systems, and insufficient consideration of long-term effects of production technologies on resource quality. However, a promising perspective for effective policy support lies in the possible link of the regional LUPAS approach with farm household models.

Diversity of Global Rice Markets and the Science Required for Consumer-Targeted Rice Breeding

With the ever-increasing global demand for high quality rice in both local production regions and with Western consumers, we have a strong desire to understand better the importance of the different traits that make up the quality of the rice grain and obtain a full picture of rice quality demographics. Rice is by no means a ‘one size fits all’ crop. Regional preferences are not only striking, they drive the market and hence are of major economic importance in any rice breeding / improvement strategy. In this analysis, we have engaged local experts across the world to perform a full assessment of all the major rice quality trait characteristics and importantly, to determine how these are combined in the most preferred varieties for each of their regions. Physical as well as biochemical characteristics have been monitored and this has resulted in the identification of no less than 18 quality trait combinations. This complexity immediately reveals the extent of the specificity of consumer preference. Nevertheless, further assessment of these combinations at the variety level reveals that several groups still comprise varieties which consumers can readily identify as being different. This emphasises the shortcomings in the current tools we have available to assess rice quality and raises the issue of how we might correct for this in the future. Only with additional tools and research will we be able to define directed strategies for rice breeding which are able to combine important agronomic features with the demands of local consumers for specific quality attributes and hence, design new, improved crop varieties which will be awarded success in the global market.

Random forest predictive modeling of mineral prospectivity with small number of prospects and data with missing values in Abra (Philippines)

Machine learning methods that have been used in data-driven predictive modeling of mineral prospectivity (e.g., artificial neural networks) invariably require large number of training prospect/locations and are unable to handle missing values in certain evidential data. The Random Forests (RF) algorithm, which is a machine learning method, has recently been applied to data-driven predictive mapping of mineral prospectivity, and so it is instructive to further study its efficacy in this particular field. This case study, carried out using data from Abra (Philippines), examines (a) if RF modeling can be used for data-driven modeling of mineral prospectivity in areas with a few (i.e., <20) mineral occurrences and (b) if RF modeling can handle evidential data with missing values. We found that RF modeling outperforms weights-of-evidence (WofE) modeling of porphyry-Cu prospectivity in the Abra area, where 12 porphyry-Cu prospects are known to exist. Moreover, just like WofE modeling, RF modeling allows analysis of the spatial associations of known prospects with individual layers of evidential data. Furthermore, RF modeling can handle missing values in evidential data through an RF-based imputation technique whereas in WofE modeling values are simply represented by zero weights. Therefore, the RF algorithm is potentially more useful than existing methods that are currently used for data-driven predictive mapping of mineral prospectivity. In particular, it is not a purely black-box method like artificial neural networks in the context of data-driven predictive modeling of mineral prospectivity. However, further testing of the method in other areas with a few mineral occurrences is needed to fully investigate its usefulness in data-driven predictive modeling of mineral prospectivity. The Random Forest (RF) algorithm is tested data-driven modeling of mineral prospectivity.The RF algorithm can be used in areas with few (i.e., <20) mineral occurrences.The RF algorithm can handle evidential data with missing values.The RF algorithm allows analysis of spatial associations of prospects with every evidence layer.

Data-driven predictive modeling of mineral prospectivity using random forests: a case study in Catanduanes Island (Philippines)

The Random Forests (RF) algorithm is a machine learning method that has recently been demonstrated as a viable technique for data-driven predictive modeling of mineral prospectivity, and thus, it is instructive to further examine its usefulness in this particular field. A case study was carried out using data from Catanduanes Island (Philippines) to investigate further (a) if RF modeling can be used for data-driven modeling of mineral prospectivity in areas with few (i.e., <20) mineral occurrences and (b) if RF modeling can handle predictor variables with missing values. We found that RF modeling outperforms evidential belief (EB) modeling of prospectivity for hydrothermal Au–Cu deposits in Catanduanes Island, where 17 hydrothermal Au–Cu prospects are known to exist. Moreover, just like EB modeling, RF modeling allows analysis of the spatial relationships between known prospects and individual layers of predictor data. Furthermore, RF modeling can handle missing values in predictor data through an RF-based imputation technique whereas in EB modeling, missing values are simply represented by maximum uncertainty. Therefore, the RF algorithm is a potentially useful method for data-driven predictive modeling of mineral prospectivity in regions with few (i.e., <20) occurrences of mineral deposits of the type sought. However, further testing of the method in other regions with few mineral occurrences is warranted to fully determine its usefulness in data-driven predictive modeling of mineral prospectivity.

Combining farm and regional level modelling for Integrated Resource Management in East and South-east Asia

Abstract Currently, in many of the highly productive lowland areas of East and South-east Asia a trend to further intensification and diversification of agricultural land use can be observed. Growing economies and urbanization also increase the claims on land and water by non-agricultural uses. As a result, decisions related to the management and planning of scarce resources become increasingly complex. Technological innovations at the field/farm level are necessary but not sufficient – changes in resource use at regional scale will also be essential. To support decision-making in such situations, we advocate a multi-scale modelling approach embedded in a sound participatory process. To this end, the Integrated Resource Management and Land use Analysis (IRMLA) Project is developing an analytical framework and methods for resource use analysis and planning, for four sites in Asia. In the envisaged multi-scale approach, integration of results from field, farm, district and provincial level analysis is based on interactive multiple goal linear programming (IMGLP), farm household modelling (FHM), production ecological concepts and participatory techniques. The approach comprises the following steps: (i) inventory/quantification of current land use systems, resource availability, management practices and policy views, (ii) analysis of alternative, innovative land use systems/technologies, (iii) exploration of the opportunities and limitations to change resource use at regional scale under alternative future scenarios, (iv) modelling decision behaviour of farmers and identification of feasible policy interventions, and (v) synthesis of results from farm to regional level for negotiation of the most promising options by a stakeholder platform. In the current paper, the operationalisation of dual-scale analysis is illustrated by the outputs (development scenarios, promising policy measures and innovative production systems) from various component models for the case study Ilocos Norte, Philippines. An approach is discussed for the integration of results from the different model components at two different decision making levels (farm and province).

Farmers' welfare, food production and the environment: a model-based assessment of the effects of new technologies in the northern Philippines.

Policy objectives of attaining food self-sufficiency and improving the well-being of subsistence farmers while protecting the environment have stimulated the development of many improved agricultural production technologies. With a choice of technologies, farm household decisions are governed not only by productivity and profitability considerations but also by factors such as available resources and their quality, family consumption preferences and attitudes towards risks, and prevailing policies. It is therefore necessary to analyse the adoption of such technologies from a whole-farm perspective. In this paper, a farm household model is used to assess possible technology adoption behaviour of farmers in Ilocos Norte Province, Philippines. Four alternative technologies were evaluated: hybrid rice production (HYR), balanced fertilization strategy (BFS), site-specific nutrient management (SSNM) and integrated pest management (IPM). Possible impacts of price policies and infrastructure improvements on technology adoption were assessed. The results show that all four alternative technologies considered are attractive to farmers, although simulations show differential adoption rates for poor, average and better-off households. IPM and HYR appear the most attractive amongst all technologies considered. In all technology simulations, relative profitability and risks, labour and capital requirements and availabilities are decisive factors in the adoption of alternative technologies. Adoption of alternative technologies would result in higher discretionary income, higher rice production and lower biocide use and nitrogen loss. Amongst policy simulations considered, availability of low-cost credit shows the largest improvements in farmer welfare for poor and average households, but its effect on simulated adoption of alternative technologies was variable. We argue that the methodology and results presented can contribute to ex ante assessments of policies targeted at stimulating technology adoption by farmers.

Spectral Signature Generalization and Expansion Can Improve the Accuracy of Satellite Image Classification

Conventional supervised classification of satellite images uses a single multi-band image and coincident ground observations to construct spectral signatures of land cover classes. We compared this approach with three alternatives that derive signatures from multiple images and time periods: (1) signature generalization: spectral signatures are derived from multiple images within one season, but perhaps from different years; (2) signature expansion: spectral signatures are created with data from images acquired during different seasons of the same year; and (3) combinations of expansion and generalization. Using data for northern Laos, we assessed the quality of these different signatures to (a) classify the images used to derive the signature, and (b) for use in temporal signature extension, i.e., applying a signature obtained from data of one or several years to images from other years. When applying signatures to the images they were derived from, signature expansion improved accuracy relative to the conventional method, and variability in accuracy declined markedly. In contrast, signature generalization did not improve classification. When applying signatures to images of other years (temporal extension), the conventional method, using a signature derived from a single image, resulted in very low classification accuracy. Signature expansion also performed poorly but multi-year signature generalization performed much better and this appears to be a promising approach in the temporal extension of spectral signatures for satellite image classification.

Rapid Assessment of Crop Status: An Application of MODIS and SAR Data to Rice Areas in Leyte, Philippines Affected by Typhoon Haiyan

Asian countries strongly depend on rice production for food security. The major rice-growing season (June to October) is highly exposed to the risk of tropical storm related damage. Unbiased and transparent approaches to assess the risk of rice crop damage are essential to support mitigation and disaster response strategies in the region. This study describes and demonstrates a method for rapid, pre-event crop status assessment. The ex-post test case is Typhoon Haiyan and its impact on the rice crop in Leyte Province in the Philippines. A synthetic aperture radar (SAR) derived rice area map was used to delineate the area at risk while crop status at the moment of typhoon landfall was estimated from specific time series analysis of Moderate Resolution Imaging Spectroradiometer (MODIS) data. A spatially explicit indicator of risk of standing crop loss was calculated as the time between estimated heading date and typhoon occurrence. Results of the analysis of pre- and post-event SAR images showed that 6500 ha were flooded in northeastern Leyte. This area was also the region most at risk to storm related crop damage due to late establishment of rice. Estimates highlight that about 700 ha of rice (71% of which was in northeastern Leyte) had not reached maturity at the time of the typhoon event and a further 8400 ha (84% of which was in northeastern Leyte) were likely to be not yet harvested. We demonstrated that the proposed approach can provide pre-event, in-season information on the status of rice and other field crops and the risk of damage posed by tropical storms.

Opportunities for expanding paddy rice production in Laos: spatial predictive modeling using Random Forest

To meet the growing demand for rice and to ease the pressure on mountain ecosystems in northern Laos, it has been proposed to reduce upland rice cultivation and to expand the area under paddy rice. We used Random Forest, a classification and decision-tree-based method, to characterize the areas currently under paddy cultivation, and to predict which areas are suitable for paddy. Topographic variables and accessibility to villages and roads were the most important predictors for the presence of paddy cultivation. There appears to be much land available that is suitable for expanding paddy areas in central and southern Laos but not in the north, where more than 40% of the rice area is on sloping land, and much less area is suitable. We conclude that expanding paddy-based rice production will be difficult in most parts of northern Laos.

RiceAtlas, a spatial database of global rice calendars and production

Knowing where, when, and how much rice is planted and harvested is crucial information for understanding the effects of policy, trade, and global and technological change on food security. We developed RiceAtlas, a spatial database on the seasonal distribution of the world’s rice production. It consists of data on rice planting and harvesting dates by growing season and estimates of monthly production for all rice-producing countries. Sources used for planting and harvesting dates include global and regional databases, national publications, online reports, and expert knowledge. Monthly production data were estimated based on annual or seasonal production statistics, and planting and harvesting dates. RiceAtlas has 2,725 spatial units. Compared with available global crop calendars, RiceAtlas is nearly ten times more spatially detailed and has nearly seven times more spatial units, with at least two seasons of calendar data, making RiceAtlas the most comprehensive and detailed spatial database on rice calendar and production.