Satellite-derived land cover maps play an important role in many applications, including monitoring of smallholder-dominated agricultural landscapes. New cloud-based computing platforms and satellite sensors offer opportunities for generating land cover maps designed to meet the spatial and temporal requirements of specific applications. Such maps can be a significant improvement compared to existing products, which tend to be coarser than 300 m, are often not representative of areas with fast-paced land use change, and have a fixed set of cover classes. Here, we present two approaches for land cover classification using the Landsat archive within Google Earth Engine. Random forest classification was performed with (1) season-based composites, where median values of individual bands and vegetation indices were generated from four years for each of four seasons, and (2) metric-based composites, where different quantiles were computed for the entire four-year period. These approaches were tested for six land cover types spanning over 18,000 locations in Zambia, with ground “truth” determined by visual inspection of high-resolution imagery from Google Earth. The methods were trained on 30% of these points and tested on the remaining 70%, and results were also compared with existing land cover products. Overall accuracies of about 89% were achieved for the season- and metric-based approaches for individual classes, with 93%and 94% accuracy for distinguishing cropland from non-cropland. For the latter task, the existing Globeland30 dataset based on Landsat had much lower accuracies (around 77% on average), as did existing cover maps at coarser resolutions. Overall, the results support the use of either season or metric-based classification approaches. Both produce better results than those obtained from previous classifiers, which supports a general paradigm shift away from dependence on standard static products and towards custom generation of on-demand cover maps designed to fulfill the needs of each specific application.

One of the greatest challenges in monitoring food security is to provide reliable crop yield information that is temporally consistent and spatially scalable. An ideal yield dataset would not only extend globally and across multiple years, but would also have enough spatial granularity to characterize productivity at the field and subfield level. Rapid increases in satellite data acquisition and platforms such as Google Earth Engine that can efficiently access and process vast archives of new and historical data offer an opportunity to map yields globally, but require efficient and robust algorithms to combine various data streams into yield estimates. We recently introduced a Scalable satellite-based Crop Yield Mapper (SCYM) that combines crop models simulations with imagery and weather data to generate 30 m resolution yield estimates without the need for ground calibration. In this study, we tested new large-scale implementations of SCYM, focusing on three regions with varying crops, field sizes and landscape heterogeneity: maize in the U.S. corn belt (390,000 km²), maize in Southern Zambia (86,000 km²), and wheat in northern India (450,000 km²). As a benchmark, we also tested a simpler empirical approach (PEAKVI) that relates yield to the peak value of a time series of spatially aggregated vegetation indices, similar to methods used in current operational monitoring. Both SCYM and PEAKVI were applied to data from all Landsat's sensors and MODIS for more than a decade in each region, and evaluated against ground-based estimates at the finest available administrative level (e.g., counties in the U.S.). We found consistently high correlations (R² ≥ 0.5) between the spatial pattern of ground- and satellite-based estimates in both U.S. maize and India wheat, with small differences between methods and source of satellite data. In the U.S., SCYM outperformed PEAKVI in tracking temporal yield variations, likely owing to its explicit consideration of weather. In India, both methods failed to track temporal yield changes, with various possible explanations discussed. In Zambia, the PEAKVI approach applied to MODIS tracked yield variations much better (R² > 0.5) than any other yield estimate, likely because the frequent cloud cover in this region confounds the other approaches. Overall, this study demonstrates successful approaches to yield estimation in each region, and illustrates the importance of distinguishing between accuracy for spatial and temporal variation. The 30 m resolution of Landsat-based SCYM does not appear to offer large benefits for tracking aggregate yields, but enables finer scale analyses than possible with the other approaches.

Global biodiesel production grew by 23% per annum between 2005 and 2015, leading to a seven-fold expansion of the sector in a single decade. Rapid development in the biodiesel sector corresponded to high crude oil prices, but since mid-2014, oil prices have fallen dramatically. This paper assesses the economic and policy factors that underpinned the expansion of biodiesel, and examines the near-term prospects for biodiesel growth under conditions of low fossil fuel prices. We show that the dramatic increase in biodiesel output would not have occurred without strong policy directives, subsidies, and trade policies designed to support agricultural interests, rural economic development, energy security, and climate targets. Given the important role of policy—and the political context within each country that shapes policy objectives, instruments, and priorities—case studies of major biodiesel producing countries are presented as a key element of our analysis. Although the narrative of biodiesel policies in most countries conveys win-win outcomes across multiple objectives, the case studies show that support of particular constituents, such as farm lobbies or energy interests, often dominates policy action and generates large social costs. Looking out to 2020, the paper highlights risks to the biodiesel industry associated with ongoing regulatory and market uncertainties.

Women empowerment (WE) is increasingly viewed as an important strategy to reduce maternal and child undernutrition,^1–3 which continues to be a major health burden in low- and middle-income countries causing 3.5 million preventable maternal and child deaths, 35% of the disease burden in children younger than 5 years, and 11% of total global disability-adjusted life years.^4,5Global data show that one of the worst affected regions is sub-Saharan Africa (SSA), where about 20% of children are malnourished.^6,7 Benin is no exception, as the prevalence of stunting, wasting, and underweight was 37%, 5%, and 17%, respectively, among children aged 6 to 59 months in the 2006 Benin Demographic and Health Survey (DHS),⁸ while 9% of women had chronic energy deficiency in the 2012 DHS.⁹ Greater rates were observed in rural areas where stunting was found in 40% of children, underweight in 19%, and wasting in 5%, while 10% of women had chronic energy deficiency.^8,9 Additionally, Beninese women and children have a limited dietary diversity score (DDS), with diets predominately composed of starchy staples with little or no animal products and few fresh fruits and vegetables.^10,11 Government, United Nation agencies, and nongovernmental organizations in Benin recognize that the state of maternal and child undernutrition requires multiple types of interventions.¹²

However, women’s low empowerment status in Benin can hinder the improvement in women’s and children’s undernutrition. Indeed, although females accounted for 47% of the economically active population in 2014,¹³ social and civil legislation is strongly influenced by tradition and customs, as women continue to be required to seek their husband’s authorization in certain areas such as family planning or health services.¹⁴ Rural women provided labor to the families’ commercial plots, were responsible for household food production and processing, and also had to work in the cooperative structures set up by the state in addition to their household tasks.¹⁴ In a more recent study of productivity differences by gender in central Benin, researchers noted that female rice farmers are particularly discriminated against with regard to access to land and equipment, resulting in significant negative impacts on their productivity and income.¹⁵ As in other areas of West Africa, women also have the responsibility of caring for children and preparing food for the household,¹⁶ but they may be vulnerable to food insecurity owing to unequal intrahousehold food distribution and their willingness to forego meals in favor of children during times of scarcity.¹⁷ Finally, no study to date has examined links between women’s empowerment and nutrition in Benin.

In addition, the evidence backing the effect of women’s empowerment on maternal and child undernutrition is inconsistent.¹⁸ Using the Women’s Empowerment in Agriculture Index (WEAI), Malapit et al¹⁹ reported positive and significant association between women’s group (WG) membership, control over income, overall empowerment, and women’s health (as measured by body mass index [BMI] and DDS) in Nepal. However, in Ghana, women’s aggregate empowerment and participation in credit decisions were positively correlated with women’s DDS, but not BMI.²⁰ Mixed findings were also observed between women’s empowerment and child anthropometry. Moestue et al²¹ found a positive association between maternal involvement in social groups and length-for-age z score of 1-year-old children, but De Silva and Harpham²²showed a negative association in 6- to 18-month-old children. Shroff et al²³ found positive association between decision-making and child weight-for-age z score (WAZ), but Begum and Sen’s²⁴ analysis of Bangladesh DHS data did not reveal any significant associations. Therefore, information about which domains of WE are associated with nutritional status is limited,²⁰ and this lack of knowledge constrains the set of policy options that can be used to empower women and improve nutrition.

In addition to a limited set of studies in SSA, examinations of the effects of WE on nutrition outcomes are constrained due to interstudy differences in population characteristics, settings, or methods/conceptualizations of WE.^25–27 For example, despite recognition of the complex, multidimensional, and culturally defined nature and influence of empowerment on nutrition,^20,26,28,29 only a few studies considered the multidimensional structure of empowerment domains in Africa or examined the varied relationships between each measure of WE and maternal and child nutrition status.^30,31 Furthermore, in 2012, the International Food Policy Research Institute developed WEAI constructed from 5 prespecified domains of empowerment,³²which may not be equally relevant in all areas. In contrast, in 2015, the United Nations adopted the Sustainable Development Goals (SDG), but the specific indicators for the SDG empowerment targets are largely equality metrics.³³ To address the need for multidimensional and contextual examinations of WE and its influence on maternal and child health outcomes, we draw from the concepts put forward in the WEAI and the SDGs but took an approach more along the lines of the World Bank which gathers indicators, both equity and empowerment related, that can be used in contextually appropriate ways.³⁴ The aims of this study were therefore to first explore the structure and domains of WE in Kalalé district of northern Benin and then to examine the effects of these constructs on nutritional status of women and their children in the region.

The gap between yield potential and average farmers’ yield measures the capacity for yield improvement with current technology. The North China Plain (NCP) is a major maize producing region of China, and improving maize yield of NCP is essential to food security of the country. Some previous studies have found a substantial maize yield gap in this region (∼100% of average yields), whereas others have reported much smaller gaps. This study used remote sensing estimated yield at 30-m resolution to quantify county level yield distributions, and then used these distributions to calculate yield gaps and the persistence level of yield for 76 counties in NCP. The average yield was 8.66 t/ha across county years, and the averaged county-level yield gap, as measured by the difference between the top 10 percentile of yields and the average yield of each county, was 0.76 t/ha, or 8.7% of the average yield. When measured as the difference between maximum and average yields in each county, the estimated gap increased to an average of 31%. We also evaluated the persistence level of farmers’ yield performance, as an indicator of how much gap might be reduced by propagating agronomic practices of the highest yielding farmers. The average of yield gap persistence was 25.9% of the average yield gap, or 2.3% of average yield with a range from 0.4% to 5.3% across counties. The distance to major rivers was identified as one factor with a significant effect on yield. Nevertheless, there was tremendous spatial heterogeneity in yield persistence level across NCP, and further analysis within individual counties is required to better prioritize means to shrink the yield gap.

The emergence of satellite sensors that can routinely observe millions of individual smallholder farms raises possibilities for monitoring and understanding agricultural productivity in many regions of the world. Here we demonstrate the potential to track smallholder maize yield variation in western Kenya, using a combination of 1-m Terra Bella imagery and intensive field sampling on thousands of fields over 2 y. We find that agreement between satellite-based and traditional field survey-based yield estimates depends significantly on the quality of the field-based measures, with agreement highest (R² up to 0.4) when using precise field measures of plot area and when using larger fields for which rounding errors are smaller. We further show that satellite-based measures are able to detect positive yield responses to fertilizer and hybrid seed inputs and that the inferred responses are statistically indistinguishable from estimates based on survey-based yields. These results suggest that high-resolution satellite imagery can be used to make predictions of smallholder agricultural productivity that are roughly as accurate as the survey-based measures traditionally used in research and policy applications, and they indicate a substantial near-term potential to quickly generate useful datasets on productivity in smallholder systems, even with minimal or no field training data. Such datasets could rapidly accelerate learning about which interventions in smallholder systems have the most positive impact, thus enabling more rapid transformation of rural livelihoods.