No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Comparing empirical and survey-based yield forecasts in a dryland agro-ecosystem
Abstract
Accurate crop yield forecasts before harvest are crucial for providing early warning of agricultural losses, so that policy-makers can take steps to minimize hunger risk. Within-season surveys of farmers’ end-of-season harvest expectations are one important method governments use to develop yield forecasts. Survey-based methods have two potential limitations whose effects are poorly understood. First, survey-based forecasts may be subject to errors and biases in the response data. For example, the weather variables that most impact yields may not be the same as those that farmers consider when shaping their yield expectations, thereby undermining forecast accuracy. Secondly, surveys are typically conducted late in the growing season, giving the government less advance notices of potential crop failures or low yields, and are costly to implement. Here we investigate these limitations within the context of Zambia's annual Crop Forecast Survey (CFS). Concerning the first limitation, we analyzed the differences between CFS-predicted yields and reported yields collected by Post Harvest Surveys, and found that excess rainfall during the planting stage was more important to the actual yield than to farmers’ yield forecasts. For the second limitation, we evaluated whether a simple empirical yield forecast model could produce earlier and more accurate yield forecasts than the CFS. A random forest model using weather variables, soil texture, and soil pH as predictors were able to produce yield forecasts at the same or higher accuracy since the planting season.
... Specifically, we used the HydroBlocks land surface model (Chaney et al., 2016;Vergopolan et al., 2020) to simulate root zone soil moisture and surface temperature at a high spatial and temporal resolution (30 m; 3 h interval) over a long duration . We combine these field-scale measures with meteorological variables, remotely sensed vegetation indices, and several other socioeconomic and physical measures with a random forest (RF) model (Breiman, 2001) to predict annual maize yields at both district and field scales for Zambia (750 000 km 2 ), a southern African country that is exposed to substantial climate variability and where much of the population still depends on small-scale agriculture (Zhao et al., 2018). We use this modeling framework to answer the following questions: ...
... -Model skill. The modeling approach that we developed was able to estimate maize yields at district scale with comparable skill (R 2 = 0.57; MAE = 310 kg ha −1 ) compared to state-of-the-art approaches based on mechanistic yield models (e.g., Jin et al., 2017;Azzari et al., 2017) and with higher skill than standard empirical approaches based on weather variables or vegetation indices (Estes et al., 2013;Zhao et al., 2018). ...
... tial distribution of rainfall. Historically, yields in Zambia are higher in locations with more precipitation during the growing season (Zhao et al., 2018). On the other hand, in the RFE analysis, monthly precipitation (i.e., dynamic) only ranked 10th. ...
Soil moisture is highly variable in space and time, and deficits (i.e., droughts) play an important role in modulating crop yields. Limited hydroclimate and yield data, however, hamper drought impact monitoring and assessment at the farm field scale. This study demonstrates the potential of using field-scale soil moisture simulations to support high-resolution agricultural yield prediction and drought monitoring at the smallholder farm field scale. We present a multiscale modeling approach that combines HydroBlocks – a physically based hyper-resolution land surface model (LSM) – with machine learning. We used HydroBlocks to simulate root zone soil moisture and soil temperature in Zambia at 3 h 30 m resolution. These simulations, along with remotely sensed vegetation indices, meteorological data, and descriptors of the physical landscape (related to topography, land cover, and soils) were combined with district-level maize data to train a random forest (RF) model to predict maize yields at district and field scales (250 m). Our model predicted yields with an average testing coefficient of determination (R2) of 0.57 and mean absolute error (MAE) of 310 kgha-1 using year-based cross-validation. Our predicted maize losses due to the 2015–2016 El Niño drought agreed well with losses reported by the Food and Agriculture Organization (FAO). Our results reveal that soil moisture is the strongest and most reliable predictor of maize yield, driving its spatial and temporal variability. Soil moisture was also a more effective indicator of drought impacts on crops than precipitation, soil and air temperatures, and remotely sensed normalized difference vegetation index (NDVI)-based drought indices. This study demonstrates how field-scale modeling can help bridge the spatial-scale gap between drought monitoring and agricultural impacts.
... As mentioned above, national and sub-national agricultural statistics (e.g. FAOSTAT) can be used to understand impacts of soil, weather, and management at broad scales (Lobell and Asner, 2003;Lobell and Field, 2007;Ben-Ari and Makowski, 2014;Chabala et al., 2015;Iizumi and Ramankutty, 2016;Zhao et al., 2018;Vergopolan et al., 2021), but lack the field-level detail needed to link specific management practices to yield outcomes. ...
... The spatial scale of the crop model simulations was the 72 pre-2013 district boundaries of Zambia. Using the pre-2013 boundaries also allowed for backward compatibility with government agricultural surveys, like the Crop Forecast Survey (CFS) and Post-Harvest Survey (PHS), and ready comparison with previous agricultural studies in Zambia (Zhao et al., 2018); (Vergopolan et al., 2021). ...
Smallholder agriculture is critical for current and future food security, yet quantifying the sources of smallholder yield variance remains a major challenge. Attributing yield variance to farmer management, as opposed to soil and weather constraints, is an important step to understanding the impact of farmer decision-making, in a context where smallholder farmers use a wide range of management practices and may have limited access to fertilizer. This study used a process-based crop model to simulate smallholder maize (Zea mays) yield at the district-level in Zambia and quantify the percent of yield variance (effect size) attributed to soil, weather, and three management inputs (cultivar, fertilizer, planting date). Effect sizes were calculated via an ANOVA variance decomposition. Further, to better understand the treatment effects of management practices, effect sizes were calculated both for all years combined and for individual years. We found that farmer management decisions explained 27-82 % of total yield variance for different agro-ecological regions in Zambia, primarily due to fertilizer impact. Fertilizer explained 45 % of yield variance for the average district, although its effect was much larger in northern districts of Zambia that typically have higher precipitation, where it explained 72 % of yield variance on average. When fixing a specific fertilizer amount, the "low-cost" management options of varying planting dates and cultivars explained 20-28 % of yield variance, with some regional variation. To better understand why management practices impact yield more in particular years, we performed a correlation analysis comparing yearly management effect sizes with four meteorologically based variables: total growing season precipitation, rainy season onset, extreme heat degree days, and longest dry spell. Results showed that fertilizer's impact generally increased under favorable weather conditions, and planting date's impact increased under adverse weather conditions. This study demonstrates how a national yield variance decomposition can be used to understand where specific management interventions would have a greater impact and can provide policymakers with quantification of soil, weather, and management effects. In addition, the variance composition can easily be adapted to a different range of management inputs, such as other cultivars or fertilizer quantities, and can also be used to assess the effect size of management adaptations under climate change.
... A traditional yield prediction method is survey-based, which is time-consuming and labor-intensive [4]. Also, some yield prediction methods are adopting physical simulation models, which require management-related input parameters that are difficult to obtain [5]. ...
... (2), (3), (4), N means the number of samples. Eq. (2) is latent loss, which controls the distribution of z. ...
In the U.S., corn is the most produced crop and has been an essential part of the American diet. To meet the demand for supply chain management and regional food security, accurate and timely large-scale corn yield prediction is attracting more attention in precision agriculture. Recently, remote sensing technology and machine learning methods have been widely explored for crop yield prediction. Currently, most county-level yield prediction models use county-level mean variables for prediction, ignoring much detailed information. Moreover, inconsistent spatial resolution between crop area and satellite sensors results in mixed pixels, which may decrease the prediction accuracy. Only a few works have addressed the mixed pixels problem in large-scale crop yield prediction. To address the information loss and mixed pixels problem, we developed a variational autoencoder (VAE) based multiple instance regression (MIR) model for large-scaled corn yield prediction. We use all unlabeled data to train a VAE and the well-trained VAE for anomaly detection. As a preprocess method, anomaly detection can help MIR find a better representation of every bag than traditional MIR methods, thus better performing in large-scale corn yield prediction. Our experiments showed that variational autoencoder based multiple instance regression (VAEMIR) outperformed all baseline methods in large-scale corn yield prediction. Though a suitable meta parameter is required, VAEMIR shows excellent potential in feature learning and extraction for large-scale corn yield prediction.
... Cross-validation (Allen, 1974;Stone, 1974) was introduced as an effective method for both model assessment and model selection when the data are relatively small. A common type of cross-validation is the leave-one-out (LOO) cross-validation that has been used in many crop models (Kogan et al., 2013;Zhao et al., 2018;Li et al., 2019). This approach relies on two datasets: a training dataset is used to calibrate the model, and a testing dataset is used to assess its quality. ...
... Dividing a very small database into such three datasets is very difficult. The LOO method has been used as a cross-validation tool to calibrate, select, and assess competing models (Kogan et al., 2013;Zhao et al., 2018). It was shown in this paper that LOO can be misleading because it uses only one dataset to choose the best model and estimate its generalization skills simultaneously. ...
The use of statistical models to study the impact of weather on crop yield has not ceased to increase. Unfortunately, this type of application is characterized by datasets with a very limited number of samples (typically one sample per year). In general, statistical inference uses three datasets: the training dataset to optimize the model parameters, the validation dataset to select the best model, and the testing dataset to evaluate the model generalization ability. Splitting the overall database into three datasets is often impossible in crop yield modelling due to the limited number of samples. The leave-one-out cross-validation method, or simply leave one out (LOO), is often used to assess model performance or to select among competing models when the sample size is small. However, the model choice is typically made using only the testing dataset, which can be misleading by favouring unnecessarily complex models. The nested cross-validation approach was introduced in machine learning to avoid this problem by truly utilizing three datasets even with limited databases. In this study, we propose one particular implementation of the nested cross-validation, called the nested leave-two-out cross-validation method or simply the leave two out (LTO), to choose the best model with an optimal model selection (using the validation dataset) and estimate the true model quality (using the testing dataset). Two applications are considered: robusta coffee in Cu M'gar (Dak Lak, Vietnam) and grain maize over 96 French departments. In both cases, LOO is misleading by choosing models that are too complex; LTO indicates that simpler models actually perform better when a reliable generalization test is considered. The simple models obtained using the LTO approach have improved yield anomaly forecasting skills in both study crops. This LTO approach can also be used in seasonal forecasting applications. We suggest that the LTO method should become a standard procedure for statistical crop modelling.
... In our analysis, we have considered the effect of precipitation on growth rates, hypothesizing that precipitation would impact agricultural yields and rural livelihoods in ways that would contribute to a rural to urban movement of people (e.g. Zhao et al 2018). However, it is doubtless complicated; migration is highly contextualized (Sen 1982). ...
... Taken together, we cautiously interpret rainfall as both a pull and push factor, recognizing intra-annual rainfall patterns are important and small but frequent deviations in rainfall, or slow rainfall onset, may be as important as larger, more acute changes (Cattaneo et al. 2019). Furthermore, this metric our use of annual averages obscures the intra-annual rainfall patterns, agriculture is influenced by-other climate measures like temperature and combined effects of temperature and moisture (Henderson et al. 2017;Zhao et al. 2018). We find that SUA's growth rates are significantly associated with their distance to the primary city (i.e. the largest city in the country) in all three time periods central place theory offers an explanation for these patterns. ...
ContextTwo-fifths of Africans reside in urban areas with populations of less than 250,000. Projections estimate that by 2050 an additional one billion people will live in urban areas, causing an acceleration of growth for these smaller urban areas. While research and development have focused on primary cities with large populations, less is known about the dynamics of urban growth in smaller, “secondary” urban areas (SUA’s).Objectives
We document the spatial distribution and temporal patterns of SUA’s in eight countries across Southern Africa between 1975 and 2015. We further explore the relationships between SUA’s growth rates and climate, land use and geographic proximity to other urban areas.Methods
Our analysis integrates spatially explicit gridded population, land use, infrastructure and climate datasets. We use descriptive statistics and spatial lag and ordinary least squares regression models to quantify SUA growth rates across three periods and explore factors that are associated with the SUA growth patterns.ResultsAverage SUA growth rates are 2.44% between 1975 and 1990. We show that the climate, distance and land use significantly relate to urbanization trajectories. In addition, we find that the proximity of SUA to the largest cities also significantly relates to urban growth.Conclusions
Our results highlight the importance of SUA’s within broader African urbanization trends. SUA are undergoing rapid population changes and are important components of economic development processes and livelihoods. Quantifying patterns of SUA urbanization is important for elevating these small but critically important urban areas into the broader context of sustainable urbanization in Africa.
... This currently does not align with the SOTA, which is why new methods to improve upon this are in huge demand. Most existing methods rely on single-image-based analyses [3] or other more empirical methods [18]. Plants however are dynamic structures and exhibit various kinds of motion, which we can exploit to gain an understanding of the mass of the harvestable crop. ...
Yield forecasting is an essential task in modern agriculture, as it enables farmers and food economists to manage crop and its distribution precisely and effectively. Traditionally, most methods for yield forecasting are based on historical data and yield estimates from manually collected samples. More modern approaches often rely on computer vision-based fruit counting algorithms, which do not take individual crop weights into account. In this paper, we propose a novel, non-destructive method to estimate the mass of individual pieces of fruit by exploiting the dynamic properties of plants. By observing short-term oscilla-tory plant motion through RGB-D video data, we formulate an approach for mass estimation based on determining the parameters of a damped harmonic oscillator model. We test the proposed algorithm by collecting a dataset of around 300 video samples of strawberries on a real strawberry farm and apply our method. With a semi-automated toolchain, capable of inferring the key parameters from video data and calculating the mass of individual berries from those, we were able to estimate the mass of all berries in our dataset with a median error of 1.16 g, outperforming a baseline utilising vision-based volume estimation to infer the mass. These insights hold valuable improvements for the development of yield forecasting systems and selective harvesters, which help to address the sustainability of food production and labour shortages.
... Accurate crop yield predictions have a significant role in supporting farmers' informed economic and managerial decisions. These predictions also contribute to global efforts aimed at preventing hunger and ensuring food security (Zhao et al., 2018). ...
This study investigates the application of Artificial Neural Networks (ANN) in forecasting agricultural yields in Kazakhstan, highlighting its implications for economic management and policy-making. Utilizing data from the Bureau of National Statistics of the Republic of Kazakhstan (2000-2023), the research develops two ANN models using the Neural Net Fitting library in MATLAB. The first model predicts the total gross yield of main agricultural crops, while the second forecasts the share of individual crops, including cereals, oilseeds, potatoes, vegetables, melons, and sugar beets. The models demonstrate high accuracy, with the total gross yield model achieving an R-squared value of 0.98 and the individual crop model showing an R value of 0.99375. These results indicate a strong predictive capability, essential for practical agricultural and economic planning. The study extends previous research by incorporating a comprehensive range of climatic and agrochemical data, enhancing the precision of yield predictions. The findings have significant implications for Kazakhstan's economy. Accurate yield predictions can optimize agricultural planning, contribute to food security, and inform policy decisions. The successful application of ANN models showcases the potential of AI and machine learning in agriculture, suggesting a pathway towards more efficient, sustainable farming practices and improved quality management systems.
... Crop yield estimation is carried out by following three main methods: (1) remote sensing [2][3][4][5][6][7], (2) UAV hyperspectral [8][9][10][11][12][13], (3) modeling of crop growth conditions [14][15][16][17][18][19]. The remote sensing method and UAV hyperspectral method use remote sensing imagery and the crop canopy spectrum to predict crop yield through vegetation index. ...
Real-time crop harvest data acquisition from harvesters during harvesting operations is an important way to understand the distribution of crop harvest in the field. Most real-time monitoring systems for grain yield using sensors are vulnerable to factors such as low accuracy and low real-time performance. To address this phenomenon, a real-time grain yield monitoring system was designed in this study. The real-time monitoring of yield was accomplished by adding three pairs of photoelectric sensors to the elevator of the corn kernel harvester. The system mainly consists of a signal acquisition and processing module, a positioning module and a visualization terminal; the signal acquisition frequency was set to 1 kHz and the response time was 2 ms. When the system operated, the signal acquisition and processing module detected the sensor signal duration of grain blocking the scrapers of the grain elevator in real-time and used the low-potential signal-based corn grain yield calculation model constructed in this study to complete the real-time yield measurement. The results of the bench tests, conducted under several different operating conditions with the simulated elevator test bench built, showed that the error of the system measurement was less than 5%. Field tests were conducted on a Zoomlion 4YZL-5BZH combined corn kernel harvester and the results showed that the average error of measured yield was 3.72%. Compared to the yield measurement method using the weighing method, the average error of the bench test yield measurement was 7.6% and the average error of yield measurement in field trials with a mass flow sensor yield measurement system was 16.38%. It was verified that the system designed in this study has high yield measurement accuracy and real-time yield measurement, and can provide reference for precision agriculture and high yield management.
... In [1] paper investigates the author discuss about Zambia's Crop Forecast Survey (CFS) and its limitations. ...
Machine learning techniques with high performance computing technologies can create various new opportunities in the agriculture domain. This paper does comprehensivereview of various papers which are concentrating on machine learning (ML) and deep learning application in agriculture. This paper is categorized into three sections a) Yield prediction using machine learning technique b) Price prediction c) Leaf disease detection using neural networks. In this paper we study the comparison of neural network models with existing models. The findings of this survey paper indicate Deep learning models give high accuracy and outperform traditional image processing technique and ML techniques outperforms various traditional techniques in prediction.
... The leave-one-out method (Kogan et al., 2013;Zhao et al., 2018;Li et al., 2019) is used here. Given N available samples (i.e., 19 years), the model is calibrated N times, using N − 1 samples (leaving one sample out). ...
Weather and climate strongly impact coffee; however, few studies have measured this impact on robusta coffee yield. This is because the yield record is not long enough, and/or the data are only available at a local farm level. A data-driven approach is developed here to 1) identify how sensitive Vietnamese robusta coffee is to weather on district and provincial levels, 2) during which key moments weather is most influential for yield, and 3) how long before harvest, yield could potentially be forecasted. Robusta coffee yield time series were available from 2000 to 2018 for the Central Highlands, where 40% of global robusta coffee is produced. Multiple linear regression has been used to assess the effect of weather on coffee yield, with regularization techniques such as PCA and leave-one-out to avoid over-fitting the regression models. The data suggest that robusta coffee in Vietnam is most sensitive to two key moments: a prolonged rainy season of the previous year favoring vegetative growth, thereby increasing the potential yield (i.e., number of fruiting nodes), while low rainfall during bean formation decreases yield. Depending on location, these moments could be used to forecast the yield anomaly with 3–6 months’ anticipation. The sensitivity of yield anomalies to weather varied substantially between provinces and even districts. In Dak Lak and some Lam Dong districts, weather explained up to 36% of the robusta coffee yield anomalies variation, while low sensitivities were identified in Dak Nong and Gia Lai districts. Our statistical model can be used as a seasonal forecasting tool for the management of coffee production. It can also be applied to climate change studies, i.e., using this statistical model in climate simulations to see the tendency of coffee in the following decades.
... Gradient boosting is one of the ensemble based supervised machine learning algorithms providing solutions to numerous agrarian applications. With respect to the application of crop yield prediction [46], gradient boosting algorithm has contributed to the construction of gridded crop models with downscaled spatial temporal approaches [11]. Analysis of the land satellite data to determine the land coverage for cropping [40], urban classification and formulating the essential vegetation indices using the gradient boosting algorithm is attained [41]. ...
The development in science and technical intelligence has incited to represent an extensive amount of data from various fields of agriculture. Therefore an objective rises up for the examination of the available data and integrating with processes like crop enhancement, yield prediction, examination of plant infections etc. Machine learning has up surged with tremendous processing techniques to perceive new contingencies in the multidisciplinary agrarian advancements. In this paper a novel hybrid regression algorithm, reinforced extreme gradient boosting is proposed which displays essentially improved execution over traditional machine learning algorithms like artificial neural networks, deep Q-Network, gradient boosting, random forest and decision tree. Extreme gradient boosting constructs new models, which are essentially, decision trees learning from the mistakes of their predecessors by optimizing the gradient descent loss function. The proposed hybrid model performs reinforcement learning at every node during the node splitting process of the decision tree construction. This leads to effective utilizationofthesamplesbyselectingtheappropriatesplitattributeforenhancedperformance. Model'sperformanceisevaluated by means of Mean Square Error, Root Mean Square Error, Mean Absolute Error, and Coefficient of Determination. To assure a fair assessment of the results, the model assessment is performed on both training and test dataset. The regression diagnostic plots from residuals and the results obtained evidently delineates the fact that proposed hybrid approach performs better with reduced error measure and improved accuracy of 94.15% over the other machine learning algorithms. Also the performance of probability density function for the proposed model delineates that, it can preserve the actual distributional characteristics of the original crop yield data more approximately when compared to the other experimented machine learning models.
... -Model skill: The modeling approach that we developed was able to estimate maize yields at district scale with comparable or higher skill (R 2 =0.61, MAE=349 kg ha −1 ) compared to state-of-the-art approaches based on mechanistic yield models (e.g., Jin et al., 2017;Azzari et al., 2017) and higher skill than standard empirical approaches based on weather variables or vegetation indices (Estes et al., 2013;Zhao et al., 2018). ...
Soil moisture is highly variable in space, and its deficits (i.e. droughts) plays an important role in modulating crop yields and its variability across landscapes. Limited hydroclimate and yield data, however, hampers drought impact monitoring and assessment at the farmer field-scale. This study demonstrates the potential of field-scale soil moisture simulations to advance high-resolution agricultural yield prediction and drought monitoring at the smallholder farm field-scale. We present a multi-scale modeling approach that combines HydroBlocks, a physically-based hyper-resolution Land Surface Model (LSM), and machine learning. We applied HydroBlocks to simulate root zone soil moisture and soil temperature in Zambia at 3-hourly 30-m resolution. These simulations along with remotely sensed vegetation indices, meteorological conditions, and data describing the physical properties of the landscape (topography, land cover, soil properties) were combined with district-level maize data to train a random forest model (RF) to predict maize yields at the district- and field-scale (250-m) levels. Our model predicted yields with a coefficient of variation (R2) of 0.61, Mean Absolute Error (MAE) of 349 kg ha−1, and mean normalized error of 22 %. We captured maize losses due to the 2015/2016 El Niño drought at similar levels to losses reported by the Food and Agriculture Organization (FAO). Our results revealed that soil moisture is the strongest and most reliable predictor of maize yield, driving its spatial and temporal variability. Consequently, soil moisture was also the most effective indicator of drought impacts in crops when compared with precipitation, soil and air temperatures, and remotely-sensed NDVI-based drought indices. By combining field-scale root zone soil moisture estimates with observed maize yield data, this research demonstrates how field-scale modeling can help bridge the spatial scale discontinuity gap between drought monitoring and agricultural impacts.
... Several methods have been used to estimate crop yields (Mavromatis, 2016;Lai et al., 2018;Zhao et al., 2018b). In this work, stepwise regression technique was used to develop a new model for estimating soft wheat yield as a function of factor costs (Riwthong et al., 2017). ...
In order to improve socio-economic conditions in rural areas, the purpose of this study is to develop models for estimating wheat yield based on the use of agricultural inputs. The emphasis is not only on increasing production, but on economically efficient production that optimises productive resources. Thereby, a cost efficiency analysis for wheat producers was conducted based on a sample of 111 farmers in the Tadla Plain in central Morocco. To achieve this objective, two models devoted to wheat yield estimation in terms of costs and quantities of agricultural inputs were developed, based on stepwise regression approach. Furthermore, data envelopment analysis (DEA) method was used to analyse the cost efficiency of farms. Overall, this work separates the excess costs related to technical efficiency and economies of scale to cost efficiency in order to understand the type of actions that farmers can take to reduce production costs and ensure their resilience.
... The proposed merging framework leverages SMAP potential by providing high-resolution and accurate soil moisture estimates that are relevant for field-scale water resources decision making. For instance, 30-m soil moisture data can improve estimates of agricultural yields and water demand at field scale (Ines et al., 2013;Fisher et al., 2017;Zhao et al., 2018;Waldman et al., 2019). If we fully trust SMAP estimates and do not bias correct the brightness temperature estimates, the 30-downscaled soil moisture can help track the large-scale impact of human activities, such as irrigation (Mathias et al., 2017;Lawston et al., 2017;Dirmeyer and Norton, 2018). ...
... Several methods have been used to estimate crop yields (Mavromatis, 2016;Lai et al., 2018;Zhao et al., 2018b). In this work, stepwise regression technique was used to develop a new model for estimating soft wheat yield as a function of factor costs (Riwthong et al., 2017). ...
Abstract: In order to improve socio-economic conditions in rural areas, the purpose of this study is to develop models for estimating wheat yield based on the use of agricultural inputs. The emphasis is not only on increasing production, but on economically efficient production that optimises productive resources. Thereby, a cost efficiency analysis for wheat producers was conducted based on a sample of 111 farmers in the Tadla Plain in central Morocco. To achieve this objective, two models devoted to wheat yield estimation in terms of costs and quantities of agricultural inputs were developed, based on stepwise regression approach. Furthermore, data envelopment analysis (DEA) method was used to analyse the cost efficiency of farms. Overall, this work separates the excess costs related to technical efficiency and economies of scale to cost efficiency in order to understand the type of actions that farmers can take to reduce production costs and ensure their resilience.
Keywords: Wheat farm; yield model; data envelopment analysis; cost efficiency; performance; profitability; Morocco.
... The main strength of the CRU dataset is that it compiles station data of multiple variables from numerous data sources into a consistent format, providing high spatial resolution and long time series, which is useful for multidecadal analysis (NCAR 2018). Recent studies around the world (e.g., Bless et al. 2018;Zhao et al. 2018;Kannenberg et al. 2018;Zambreski et al. 2018;Liu et al. 2018) made use of this data. A modified version of the Penman-Monteith reference evapotranspiration (ETo) equation is used in the CRU dataset as a metric of the AED. ...
Droughts are complex and may be triggered by different mechanisms, such as atmospheric circulation, moisture transport, and thermodynamic processes. Significant research has been completed to characterize precipitation in the Intergovernmental Panel on Climate Change (IPCC) reference regions (RRs), but a systematic analysis of atmospheric transport linked to drought episodes is still missing. This article describes a catalog in which the drought episodes over the RRs are identified during 1980–2015, and the role of the moisture transport anomalies from the respective major climatological moisture sources during the most severe meteorological drought episode registered for each RR is analyzed. For each of the 27 RRs defined in the IPCC Fifth Assessment Report, drought episodes were identified at 1-, 6-, and 12-month time scales through the standardized precipitation evapotranspiration index (SPEI). SPEI values were computed using time series of the monthly precipitation and atmospheric evaporative demand (AED) averaged over each RR. The approach, which was applied to both identify the major climatological moisture sources and sinks for each RR and to investigate anomalies in moisture transport during the episode, is based on the Lagrangian flexible particle dispersion model (FLEXPART), integrated with the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim) data. For each RR, the following components were analyzed: a) moisture uptake over sources, b) moisture supply from the sources into the RR, and c) moisture supply from the RR into its sink. Although performed for just one case, this analysis illustrates how the moisture transport may impact the RR during extreme conditions. The results are organized in a web page available to the scientific community and stakeholders.
Corn occupies a significant portion of the American residents’ diet. Therefore, accurate prediction of corn’s annual yield in agricultural cultivation would greatly assist farmers in improving planting efficiency and have significant implications for agricultural resource management, market planning, food safety monitoring, and other related fields. To enhance the accuracy of corn yield prediction, this research utilized remote sensing satellite data and employed the Multiple Instance Learning and Dictionary Learning (MIDL) method to predict county-level corn yield within 12 states in the Midwest region of the United States. MIDL combines Multiple Instance Learning to retain detailed information within each county and Dictionary Learning to filter and eliminate potentially interfering mixed pixels’ information in the prediction process. Experimental results demonstrate that MIDL achieved high prediction accuracy and exhibited excellent spatial generalization capabilities, outperforming all the compared methods. This study extensively analyzed the strengths and weaknesses of MIDL, confirming its promising potential, and identified directions for future improvements. Proposed MIDL method had the best performance for the six testing years 2016-2021, achieving an average R
2
of 0.79.
Due to current global population growth, resource shortages, and climate change, traditional agricultural models face major challenges. Precision agriculture (PA), as a way to realize the accurate management and decision support of agricultural production processes using modern information technology, is becoming an effective method of solving these challenges. In particular, the combination of remote sensing technology and machine learning algorithms brings new possibilities for PA. However, there are relatively few comprehensive and systematic reviews on the integrated application of these two technologies. For this reason, this study conducts a systematic literature search using the Web of Science, Scopus, Google Scholar, and PubMed databases and analyzes the integrated application of remote sensing technology and machine learning algorithms in PA over the last 10 years. The study found that: (1) because of their varied characteristics, different types of remote sensing data exhibit significant differences in meeting the needs of PA, in which hyperspectral remote sensing is the most widely used method, accounting for more than 30% of the results. The application of UAV remote sensing offers the greatest potential, accounting for about 24% of data, and showing an upward trend. (2) Machine learning algorithms displays obvious advantages in promoting the development of PA, in which the support vector machine algorithm is the most widely used method, accounting for more than 20%, followed by random forest algorithm, accounting for about 18% of the methods used. In addition, this study also discusses the main challenges faced currently, such as the difficult problems regarding the acquisition and processing of high-quality remote sensing data, model interpretation, and generalization ability, and considers future development trends, such as promoting agricultural intelligence and automation, strengthening international cooperation and sharing, and the sustainable transformation of achievements. In summary, this study can provide new ideas and references for remote sensing combined with machine learning to promote the development of PA.
This rapid climate risk assessment for the Southern Africa Development Community (SADC) uses the Intergovernmental Panel on Climate Change (IPCC) 2014 risk analysis framework to assess the distribution of climate hazards and social and biophysical vulnerability to those hazards in order to identify climate risk hotspots. The assessment uses regional climate models from CORDEX-Africa to map rainfall extremes and drought hazards to 2031–2059. Ten social and biophysical vulnerability indicators are identified from across the capital assets (human, physical, social, financial, natural), using data from the Global Multidimensional Poverty Index (MPI), to develop a vulnerability index. The vulnerability index and distribution of climate hazards are mapped to identify hotspots. Hotspots of vulnerability to and risk of extreme rainfall are shown in northern Madagascar and in south west Tanzania, under both the RCP4.5 and 8.5 scenarios. These hotspots also correspond to the hotspots for drought risk under RCP4.5 and 8.5. However, it is clear that medium-high climate risk (high vulnerability, medium-high climate hazard) is widespread across Angola, Democratic Republic of the Congo (DRC), Tanzania, Mozambique, and Madagascar.
This book contains original, peer-reviewed research articles from the Second International Conference on Recent Trends in Machine Learning, IoT, Smart Cities and Applications, held in March 2021 at CMR Institute of Technology, Hyderabad, Telangana India. It covers the latest research trends and developments in areas of machine learning, artificial intelligence, neural networks, cyber-physical systems, cybernetics, with emphasis on applications in smart cities, Internet of Things, practical data science and cognition. The book focuses on the comprehensive tenets of artificial intelligence, machine learning and deep learning to emphasize its use in modelling, identification, optimization, prediction, forecasting and control of future intelligent systems. Submissions were solicited of unpublished material, and present in-depth fundamental research contributions from a methodological/application perspective in understanding artificial intelligence and machine learning approaches and their capabilities in solving a diverse range of problems in industries and its real-world applications.
The use of statistical models to study the impact of weather on crop yield has not ceased to increase. Unfortunately, this type of application is characterised by datasets with a very limited number of samples (typically one sample per year). In general, statistical inference uses three datasets: the training dataset to optimise the model parameters, the validation datasets to select the best model, and the testing dataset to evaluate the model generalisation ability. Splitting the overall database into three datasets is impossible in crop yield modelling. The leave-one-out cross-validation method or simply leave-one-out (LOO) has been introduced to facilitate statistical modelling when the database is limited. However, the model choice is made using the testing dataset, which can be misleading by favouring unnecessarily complex models. The nested cross-validation approach was introduced in machine learning to avoid this problem by truly utilising three datasets, especially problems with limited databases. In this study, we proposed one particular implementation of the nested cross-validation, called the leave-two-out method (LTO), to chose the best model with an optimal model complexity (using the validation dataset) and estimated the true model quality (using the testing dataset). Two applications are considered: Robusta coffee in Cu M'gar (Dak Lak, Vietnam) and grain maize over 96 French departments. In both cases, LOO is misleading by choosing too complex models; LTO indicates that simpler models actually perform better when a reliable generalisation test is considered. The simple models obtained using the LTO approach have reasonable yield anomaly forecasting skills in both study crops. This LTO approach can also be used in seasonal forecasting applications. We suggest that the LTO method should become a standard procedure for statistical crop modelling.
This rapid climate risk assessment for the Southern Africa Development Community (SADC) uses the Intergovernmental Panel on Climate Change (IPCC) 2014 risk analysis framework to assess the distribution of climate hazards and social and biophysical vulnerability to those hazards in order to identify climate risk hotspots.
The assessment uses regional climate models from CORDEX-Africa to map rainfall extremes and drought hazards to 2031–2059. Ten social and biophysical vulnerability indicators are identified from across the capital assets (human, physical, social, financial, natural), using data from the Global Multidimensional Poverty Index (MPI), to develop a vulnerability index. The vulnerability index and distribution of climate hazards are mapped to identify hotspots.
Hotspots of vulnerability to and risk of extreme rainfall are shown in northern Madagascar and in south west Tanzania, under both the RCP4.5 and 8.5 scenarios. These hotspots also correspond to the hotspots for drought risk under RCP4.5 and 8.5. However, it is clear that medium-high climate risk (high vulnerability, medium-high climate hazard) is widespread across Angola, Democratic Republic of the Congo (DRC), Tanzania, Mozambique, and Madagascar.
(Published by: CGIAR Research Program on Climate Change, Agriculture and Food Security)
Grain production is an essential part of the Chinese economic system. It not only related to the survival and health of each people but also plays a critical role in social stability. However, more and more foods are wasted in different stages such as harvest, production, storage, and consumption. Therefore, it is essential to accurately evaluate the loss rate of grain during different stages and deal with it accordingly. With the advantages of the information technologies, a large volume of data has been collected in the different stages of grain production, such as harvest, processing, transportation, and consumption. In this paper, we propose an integrated structure to combine multiple clustering models to analyze the grain loss rate in different stages. kNN, softmax regression, decision tree, and XGBoost algorithms are studied and integrated with the proposed combined framework. The experimental results on the survey dataset suggested that the relevant algorithms of machine learning can be combined to improve the prediction accuracy of the grain loss rates. The evaluation results indicate that the proposed method can improve the accuracy to 94% in the test dataset which is higher than any other compared methods
The purpose of this paper is to develop a maintenance decision support system (DSS) framework using in-service lubricant data for fault diagnosis. The DSS reveals embedded patterns in the data (knowledge discovery), automatically quantifies the influence of lubricant parameters on the unhealthy state of the machine using alternative classifiers. The classifiers are compared for robustness from which decision-makers select an appropriate classifier given a specific lubricant data structure.
The DSS embeds a framework integrating cluster and principal component analysis, for feature extraction, and eight classifiers among them extreme gradient boosting (XGB), random forest (RF), decision trees (DT) and logistic regression (LR). A qualitative and quantitative criterion is developed in conjunction with practitioners for comparing the classifier models.
The results show the importance of embedded knowledge, explored via a knowledge discovery approach. Moreover, the efficacy of the embedded knowledge on maintenance DSS is emphasized. Importantly, the proposed framework is demonstrated as plausible for decision support due to its high accuracy and consideration of practitioners needs.
The proposed framework will potentially assist maintenance managers in accurately exploiting lubricant data for maintenance DSS, while offering insights with reduced time and errors.
Advances in lubricant-based intelligent approach for fault diagnosis is seldom utilized in practice, however, may be incorporated in the information management systems offering high predictive accuracy. The classification models’ comparison approach, will inevitably assist the industry in selecting amongst divergent models’ for DSS.
p>Understanding the cross-scale linkages between drought and food security is vital to developing tools to reduce drought impacts and support decision making. This study reviews how drought hazards transfer to food insecurity through changes in physical processes and socio-environmental systems across a wide range of spatial and temporal scales. We propose a multi-scale, integrated framework leveraging modeling advances (e.g. drought and crop monitoring, water-food-energy nexus, decision making) and increased data availability (e.g. satellite remote sensing, food trade) through the lens of the coupled human–natural system to support multidisciplinary approaches and avoid potential policy spillover effects. We discuss current scale-dependent challenges in tackling drought-induced food security whilst minimizing water use conflicts and environmental impacts.</p
The purpose of this article is to show the extreme temperature regime of heat waves across Africa over recent years (1981–2015). Heat waves have been quantified using the Heat Wave Magnitude Index daily (HWMId), which merges the duration and the intensity of extreme temperature events into a single numerical index. The HWMId enables a comparison between heat waves with different timing and location, and it has been applied to maximum and minimum temperature records. The time series used in this study have been derived from (1) observations from the Global Summary of the Day (GSOD) and (2) reanalysis data from ERA-Interim. The analysis shows an increasing number of heat waves of both maxima and minima temperatures in the last decades. Results from heat wave analysis of maximum temperature (HWMIdtx) indicate an increase in intensity and frequency of extreme events. Specifically, from 1996 onwards it is possible to observe HWMIdtx spread with the maximum presence during 2006–2015. Between 2006 and 2015 the frequency (spatial coverage) of extreme heat waves had increased to 24.5 observations per year (60.1 % of land cover), as compared to 12.3 per year (37.3 % of land area) in the period from 1981 to 2005 for GSOD stations (reanalysis).
Vegetation plays a key role in the global climate system via modification of the water and energy balance. Its coupling to climate is therefore important particularly in the tropics, where severe climate change impacts are expected. Vegetation growth is mutually controlled by temperature and water availability while it modifies regional climate through latent heat flux and changes in albedo. Consequently, understanding how projected climate change will impact vegetation and the forcing of vegetation on climate for various land cover types in East Africa is vital. This study provides an assessment of the vegetation trends in East Africa using Leaf Area Index (LAI) time series for the period 1982 to 2011, lead/lag correlation analysis between LAI and climate, a statistical estimation of vegetation feedback on climate using lagged covariance ratios as well as spatial regression analysis. Our results show few significant changes in current LAI trends though persistent declining vegetation trends are shown from Southern Ethiopia extending through Central Kenya into Central Tanzania. Precipitation (temperature) exerts widespread positive (negative) forcing on lagging vegetation except in forests. Positive correlations between the lagging Antecedent Precipitation Index (API) and LAI were dominant compared to temperature. Positive vegetation feedback on precipitation dominates across the region while a stronger negative forcing is exerted on Tmin compared to Tmax. Spatial dependence was also shown as a key component in the vegetation-climate interactions in the region. Given the vital role of land surface dynamics on local and regional climate, these results provide a valuable point of reference for evaluating the land-atmosphere coupling in the region.
The purpose of this article is to show the extreme temperature regime of heat waves across Africa over recent years (1981–2015). Heat waves have been quantified using the Heat Wave Magnitude Index daily (HWMId), which merges the duration and the intensity of extreme temperature events into a single numerical index. The HWMId enables a comparison between heat waves with different timing and location, and it has been applied to maximum and minimum temperature records. The time series used in this study have been derived from: (1) observations from the Global Summary of the Day (GSOD); and (2) reanalysis data from ERA-INTERIM. The analysis show an increasing numbers of heat waves of both maxima and minima temperatures in the last decades. Results from heat wave analysis of maximum temperature (HWMIdtx) indicate an increase in intensity and frequency of extreme events. Specifically, from 1996 onwards it is possible to observe HWMIdtx spread with the maximum presence during 2006–2015. Between 2006 and 2015 the frequency (spatial coverage) of extreme heat waves had increased to 24.5 observations (60.1 % of land cover) per year, as compared to 12.3 (37.3 % of land area) per year in the period from 1981 to 2005 for GSOD stations (reanalysis).
The Climate Hazards group Infrared Precipitation with Stations (CHIRPS) dataset builds on previous approaches to ‘smart’ interpolation techniques and high resolution, long period of record precipitation estimates based on infrared Cold Cloud Duration (CCD) observations. The algorithm i) is built around a 0.05° climatology that incorporates satellite information to represent sparsely gauged locations, ii) incorporates daily, pentadal, and monthly 1981-present 0.05° CCD-based precipitation estimates, iii) blends station data to produce a preliminary information product with a latency of about 2 days and a final product with an average latency of about 3 weeks, and iv) uses a novel blending procedure incorporating the spatial correlation structure of CCD-estimates to assign interpolation weights. We present the CHIRPS algorithm, global and regional validation results, and show how CHIRPS can be used to quantify the hydrologic impacts of decreasing precipitation and rising air temperatures in the Greater Horn of Africa. Using the Variable Infiltration Capacity model, we show that CHIRPS can support effective hydrologic forecasts and trend analyses in southeastern Ethiopia.
This paper provides a first overview of the performance of state-of-the-art global climate models participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5) in simulating climate extremes indices defined by the Expert Team on Climate Change Detection and Indices (ETCCDI), and compares it to that in the previous model generation (CMIP3). For the first time, the indices based on daily temperature and precipitation are calculated with a consistent methodology across multimodel simulations and four reanalysis data sets (ERA40, ERA-Interim, NCEP/NCAR, and NCEP-DOE) and are made available at the ETCCDI indices archive website. Our analyses show that the CMIP5 models are generally able to simulate climate extremes and their trend patterns as represented by the indices in comparison to a gridded observational indices data set (HadEX2). The spread amongst CMIP5 models for several temperature indices is reduced compared to CMIP3 models, despite the larger number of models participating in CMIP5. Some improvements in the CMIP5 ensemble relative to CMIP3 are also found in the representation of the magnitude of precipitation indices. We find substantial discrepancies between the reanalyses, indicating considerable uncertainties regarding their simulation of extremes. The overall performance of individual models is summarized by a "portrait" diagram based on root-mean-square errors of model climatologies for each index and model relative to four reanalyses. This metric analysis shows that the median model climatology outperforms individual models for all indices, but the uncertainties related to the underlying reference data sets are reflected in the individual model performance metrics. Key PointsWe calculate indices in a consistent manner across models and reanalysesMulti-model ensembles compare reasonably well with observation-based indicesThere are large uncertainties in the representation of extremes in reanalyses
Getting farmers to adopt new cultivars with greater tolerance for coping with climatic extremes and variability is considered as one way of adapting agricultural production to climate change. However, for successful adaptation to occur, an accurate recognition and understanding of the climate signal by key stakeholders (farmers, seed suppliers and agricultural extension services) is an essential precursor. This paper presents evidence based on fieldwork with smallholder maize producers and national seed network stakeholders in Malawi from 2010 to 2011, assessing understandings of rainfall changes and decision-making about maize cultivar choices. Our findings show that preferences for short-season maize cultivars are increasing based on perceptions that season lengths are growing shorter due to climate change and the assumption that growing shorter-season crops represents a good strategy for adapting to drought. However, meteorological records for the two study areas present no evidence for shortening seasons (or any significant change to rainfall characteristics), suggesting that short-season cultivars may not be the most suitable adaptation option for these areas. This demonstrates the dangers of oversimplified climate information in guiding changes in farmer decision-making about cultivar choice.
Do the wet savannahs and shrublands of Africa provide a large reserve of potential croplands to produce food staples or bioenergy with low carbon and biodiversity costs? We find that only small percentages of these lands have meaningful potential to be low-carbon sources of maize (�2%) or soybeans (9.5–11.5%), meaning that their conversion would release at least one third less carbon per ton of crop than released on average for the production of those crops on existing croplands. Factoring in land-use change, less than 1% is likely to produce cellulosic ethanol that would meet European standards for greenhouse gas reductions. Biodiversity e�ffects of converting these lands are also likely to be significant as bird and mammal richness is comparable to that of the world’s tropical forest regions. Our findings contrast with influential studies that assume these lands provide a large, low-environmental-cost cropland reserve.
Drought is one of the leading impediments to development in Africa. Much of the continent is dependent on rain-fed agriculture, which makes it particularly susceptible to climate variability. Monitoring drought and providing timely seasonal forecasts are essential for integrated drought risk reduction. Current approaches in developing regions have generally been limited, however, in part because of unreliable monitoring networks. Operational seasonal climate forecasts are also deficient and often reliant on statistical regressions, which are unable to provide detailed information relevant for drought assessment. However, the wealth of data from satellites and recent advancements in large-scale hydrological modeling and seasonal climate model predictions have enabled the development of state-of-the-art monitoring and prediction systems that can help address many of the problems inherent to developing regions. An experimental drought monitoring and forecast system for sub-Saharan Africa is described that is based on advanced land surface modeling driven by satellite and atmospheric model data. Key elements of the system are the provision of near-real-time evaluations of the terrestrial water cycle and an assessment of drought conditions. The predictive element takes downscaled ensemble dynamical climate forecasts and provides, when merged with the hydrological modeling, ensemble hydrological forecasts. We evaluate the overall skill of the system for monitoring and predicting the development of drought and illustrate the use of the system for the 2010/11 Horn of Africa drought. A key element is the transition and testing of the technology for operational usage by African collaborators and we discuss this for two implementations in West and East Africa.
This paper investigates farmers’ perceptions of climate change and variability in southwest Uganda and compares them with daily rainfall and temperature measurements from the 1960s to the present, including trends in daily rainfall and temperature, seasonality, changing probability of risk and intensity of rainfall events. Statistical analyses and modelling of rainfall and temperature were performed and contrasted with qualitative data collected through a semi-structured questionnaire. The fieldwork showed that farmers perceived regional climate to have changed in the past 20 years. In particular, farmers felt that temperature had increased and seasonality and variability had changed, with the first rainy season between March and May becoming more variable. Farmers reported detailed accounts of climate characteristics during specific years, with recent droughts in the late 1990s and late 2000s confirming local perceptions that there has been a shift in climate towards more variable conditions that are less favourable to production. There is a clear signal that temperature has been increasing in the climate data and, to a lesser extent, evidence that the reliability of rains in the first season has decreased slightly. However, rainfall measurements do not show a downward trend in rainfall amount, a significant shift in the intensity of rainfall events or in the start and end of the rainy seasons. We explore why there are some differences between farmers’ perceptions and the climate data due to different associations of risk between ideal rainfall by farmers, including the amount and distribution needed for production, meteorological definitions of normal rainfall or the long-term statistical mean and its variation, and the impact of higher temperatures. The paper reflects on the methodological approach and considers the implications for communicating information about risk to users in order to support agricultural innovation.
Development goals and poverty-reduction policies are often focused on raising agricultural productivity and dependent on farm
household level data. Historically, household surveys commonly employed self-reported land area measurements for cost-effectiveness
and convenience. However, as we illustrate here, these self-reported estimates may measure land with systematic error resulting
in sizable biases. This has led to the increased use of Global Positioning Systems (GPS) and other modern technologies to
measure land size. In this article, we compare self-reported (SR) and GPS land measurement to assess the differences between
the measures, to identify the sources of differences, and to determine the implications of the different measures on agricultural
analysis. The results from the analysis of data from four African countries indicate that SR land areas systematically differ
from GPS land measures and that this difference leads to biased estimates of the relationship between land and productivity
and consistently low estimates of land inequality. Through the evidence and analysis presented here, we conclude that the
more systematic use of GPS-measured land area will result in improved agricultural statistics and more accurate analysis of
agricultural relationships, which will better inform future policy.
Statistical studies of rainfed maize yields in the United States and
elsewhere have indicated two clear features: a strong negative yield
response to accumulation of temperatures above 30°C (or extreme
degree days (EDD)), and a relatively weak response to seasonal rainfall.
Here we show that the process-based Agricultural Production Systems
Simulator (APSIM) is able to reproduce both of these relationships in
the Midwestern United States and provide insight into underlying
mechanisms. The predominant effects of EDD in APSIM are associated with
increased vapour pressure deficit, which contributes to water stress in
two ways: by increasing demand for soil water to sustain a given rate of
carbon assimilation, and by reducing future supply of soil water by
raising transpiration rates. APSIM computes daily water stress as the
ratio of water supply to demand, and during the critical month of July
this ratio is three times more responsive to 2°C warming than to a
20% precipitation reduction. The results suggest a relatively minor role
for direct heat stress on reproductive organs at present temperatures in
this region. Effects of elevated CO2 on transpiration
efficiency should reduce yield sensitivity to EDD in the coming decades,
but at most by 25%.
This paper describes the construction of an updated gridded climate dataset (referred to as CRU TS3.10) from monthly observations at meteorological stations across the world's land areas. Station anomalies (from 1961 to 1990 means) were interpolated into 0.5° latitude/longitude grid cells covering the global land surface (excluding Antarctica), and combined with an existing climatology to obtain absolute monthly values. The dataset includes six mostly independent climate variables (mean temperature, diurnal temperature range, precipitation, wet-day frequency, vapour pressure and cloud cover). Maximum and minimum temperatures have been arithmetically derived from these. Secondary variables (frost day frequency and potential evapotranspiration) have been estimated from the six primary variables using well-known formulae. Time series for hemispheric averages and 20 large sub-continental scale regions were calculated (for mean, maximum and minimum temperature and precipitation totals) and compared to a number of similar gridded products. The new dataset compares very favourably, with the major deviations mostly in regions and/or time periods with sparser observational data. CRU TS3.10 includes diagnostics associated with each interpolated value that indicates the number of stations used in the interpolation, allowing determination of the reliability of values in an objective way. This gridded product will be publicly available, including the input station series (http://www.cru.uea.ac.uk/ and http://badc.nerc.ac.uk/data/cru/). © 2013 Royal Meteorological Society
Understanding farmers’ perceptions of how rainfall fluctuates and changes is crucial in anticipating the impacts of changing climate patterns, as only when a problem is perceived will appropriate steps be taken to adapt to it. This article seeks to: (1) identify southern African farmers’ perceptions of rainfall, rainfall variations, and changes; (2) examine the nature of meteorological evidence for the perceived rainfall variability and change; (3) document farmers’ responses to rainfall variability; and (4) discuss why discrepancies may occur between farmers’ perceptions and meteorological observations of rainfall. Semi-structured interviews were used to identify farmers’ perceptions of rainfall changes in Botswana and Malawi. Resulting perceptions were examined in conjunction with meteorological data to assess perceived and actual rainfall with regards to: what was changing (onset, duration or cessation), and how it was changing (amount, frequency, intensity or inter-annual variability). Most farmers perceived that the rains used to start earlier and end later. Meteorological data provided no evidence to support farmer perceptions of rainfall starting as early as September (south Malawi) or October (Botswana); however, a high inter-annual variability in the timing of the onset was observed alongside an increasing number of dry days and declining amounts of rainfall at the onset and cessation of precipitation. While some rainfall patterns are associated with El Niño-Southern Oscillation (ENSO) fluctuations and larger-scale changes, one explanation for the differences between farmer perceptions and meteorological evidence is that rainfall changes can be easily confused with changes in farming system sensitivity. Our findings suggest that scientists, policymakers, and developers of climate adaptation projects need to be more in tune with farmers' and extension workers’ understandings of how weather is changing in order to improve adaptation policy formulation and implementation.
Distributed irrigation systems are those in which the water access (via pump or human power), distribution (via furrow, watering can, sprinkler, drip lines, etc.), and use all occur at or near the same location. Distributed systems are typically privately owned and managed by individuals or groups, in contrast to centralized irrigation systems, which tend to be publicly operated and involve large water extractions and distribution over significant distances for use by scores of farmers. Here we draw on a growing body of evidence on smallholder farmers, distributed irrigation systems, and land and water resource availability across sub-Saharan Africa (SSA) to show how investments in distributed smallholder irrigation technologies might be used to (i) use the water sources of SSA more productively, (ii) improve nutritional outcomes and rural development throughout SSA, and (iii) narrow the income disparities that permit widespread hunger to persist despite aggregate economic advancement.
Improved understanding of the factors that limit crop yields in farmers' fields will play an important role in increasing regional food production while minimizing environmental impacts. However, causes of spatial variability in crop yields are poorly known in many regions because of limited data availability and analysis methods. In this study, we assessed sources of between-field wheat (Triticum aestivum L.) yield variability for two growing seasons in the Yaqui Valley, Mexico. Field surveys conducted in 2001 and 2003 provided data on management practices for 68 and 80 wheat fields throughout the Valley, respectively, while yields on these fields were estimated using concurrent Landsat satellite imagery. Management-yield relationships were analyzed with t tests, linear regression, and regression trees, all of which revealed significant but year-dependent impacts of management on yields. In 2001, an unusually cool year that favored high yields, N fertilizer was the most important source of between-field variability. In 2003, a warmer year with reduced irrigation water allocations, the timing of the first postplanting irrigation was found to be the most important control. Management explained at least 50% of spatial yield variability in both years. Regression tree models, which were able to capture important nonlinearities and interactions, were more appropriate for analyzing yield controls than traditional linear models. The results of this study indicate that adjustments in management can significantly improve wheat production in the Yaqui Valley but that the relevant controls change from year to year.
Drought is one of the most frequent climate-related disasters occurring across large portions of the African continent, often with devastating consequences for the food security of agricultural households. This study proposes a novel method for calculating the empirical probability of having a significant proportion of the total agricultural area affected by drought at sub-national level. First, we used the per-pixel Vegetation Health Index (VHI) from the Advanced Very High Resolution Radiometer (AVHRR) averaged over the crop season as main drought indicator. A phenological model based on NDVI was employed for defining the start of season (SOS) and end of the grain filling stage (GFS) dates. Second, the per-pixel average VHI was aggregated for agricultural areas at sub-national level in order to obtain a drought intensity indicator. Seasonal VHI averaging according to the phenological model proved to be a valid drought indicator for the African continent, and is highly correlated with the drought events recorded during the period (1981–2009). The final results express the empirical probability of drought occurrence over both the temporal and the spatial domain, representing a promising tool for future drought monitoring.
The National Agricultural Statistics Service (NASS) of the US Department of Agriculture (USDA) produces the Cropland Data Layer (CDL) product, which is a raster-formatted, geo-referenced, crop-specific, land cover map. CDL program inputs include medium resolution satellite imagery, USDA collected ground truth and other ancillary data, such as the National Land Cover Data set. A decision tree-supervised classification method is used to generate the freely available state-level crop cover classifications and provide crop acreage estimates based upon the CDL and NASS June Agricultural Survey ground truth to the NASS Agricultural Statistics Board. This paper provides an overview of the NASS CDL program. It describes various input data, processing procedures, classification and validation, accuracy assessment, CDL product specifications, dissemination venues and the crop acreage estimation methodology. In general, total crop mapping accuracies for the 2009 CDLs ranged from 85% to 95% for the major crop categories.
There is widespread interest in the impacts of climate change on agriculture in Sub-Saharan Africa (SSA), and on the most effective investments to assist adaptation to these changes, yet the scientific basis for estimating production risks and prioritizing investments has been quite limited. Here we show that by combining historical crop production and weather data into a panel analysis, a robust model of yield response to climate change emerges for several key African crops. By mid-century, the mean estimates of aggregate production changes in SSA under our preferred model specification are − 22, − 17, − 17, − 18, and − 8% for maize, sorghum, millet, groundnut, and cassava, respectively. In all cases except cassava, there is a 95% probability that damages exceed 7%, and a 5% probability that they exceed 27%. Moreover, countries with the highest average yields have the largest projected yield losses, suggesting that well-fertilized modern seed varieties are more susceptible to heat related losses.
Changes in the global production of major crops are important drivers of food prices, food security and land use decisions. Average global yields for these commodities are determined by the performance of crops in millions of fields distributed across a range of management, soil and climate regimes. Despite the complexity of global food supply, here we show that simple measures of growing season temperatures and precipitation—spatial averages based on the locations of each crop—explain ~30% or more of year-to-year variations in global average yields for the world's six most widely grown crops. For wheat, maize and barley, there is a clearly negative response of global yields to increased temperatures. Based on these sensitivities and observed climate trends, we estimate that warming since 1981 has resulted in annual combined losses of these three crops representing roughly 40 Mt or $5 billion per year, as of 2002. While these impacts are small relative to the technological yield gains over the same period, the results demonstrate already occurring negative impacts of climate trends on crop yields at the global scale.
Agricultural and Forest Meteorology j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / a g r f o r m e t a b s t r a c t Improved understanding of the influence of climate on agricultural production is needed to cope with expected changes in temperature and precipitation, and an increasing number of undernourished peo-ple in food insecure regions. Many studies have shown the importance of seasonal climatic means in explaining crop yields. However, climate variability is expected to increase in some regions and have significant consequences on food production beyond the impacts of changes in climatic means. Here, we examined the relationship between seasonal climate and crop yields in Tanzania, focusing on maize, sorghum and rice. The impacts of both seasonal means and variability on yields were measured at the subnational scale using various statistical methods and climate data. The results indicate that both intra-and interseasonal changes in temperature and precipitation influence cereal yields in Tanzania. Seasonal temperature increases have the most important impact on yields. This study shows that in Tanzania, by 2050, projected seasonal temperature increases by 2 • C reduce average maize, sorghum, and rice yields by 13%, 8.8%, and 7.6% respectively. Potential changes in seasonal total precipitation as well as intra-seasonal temperature and precipitation variability may also impact crop yields by 2050, albeit to a lesser extent. A 20% increase in intra-seasonal precipitation variability reduces agricultural yields by 4.2%, 7.2%, and 7.6% respectively for maize, sorghum, and rice. Using our preferred model, we show that we underestimate the climatic impacts by 2050 on crop yields in Tanzania by 3.6%, 8.9%, and 28.6% for maize, sorghum and rice respectively if we focus only on climatic means and ignore climate variability. This study highlights that, in addition to shifts in growing season means, changes in intra-seasonal variability of weather may be important for future yields in Tanzania. Additionally, we argue for a need to invest in improving the climate records in these regions to enhance our understanding of these relationships.
This study compares responses to seasonal climate forecasts conducted by farmers of three agro-ecological zones of Burkina
Faso, including some who had attended local level workshops and others who had not attended the workshops. While local inequalities
and social tensions contributed to excluding some groups, about two-thirds of non-participants interviewed received the forecast
from the participants or through various means deployed by the project. Interviews revealed that almost all those who received
the forecasts by some mechanism (workshop or other) shared them with others. The data show that participants were more likely
to understand the probabilistic aspect of the forecasts and their limitations, to use the information in making management
decisions and by a wider range of responses. These differences are shown to be statistically significant. Farmers evaluated
the forecasts as accurate and useful in terms of both material and non-material considerations. These findings support the
hypothesis that participatory workshops can play a positive role in the provision of effective climate services to African
rural producers. However, this role must be assessed in the context of local dynamics of power, which shape information flows
and response options. Participation must also be understood beyond single events (such as workshops) and be grounded in sustained
interaction and commitments among stakeholders. The conclusion of this study point to lessons learned and critical insights
on the role of participation in climate-based decision support systems for rural African communities.
Soil, landscape and hybrid factors are known to influence yield and quality of corn (Zea mays L.). This study employed artificial neural network (ANN) analysis to evaluate the relative importance of selected soil, landscape
and seed hybrid factors on yield and grain quality in two Illinois, USA fields. About 7 to 13 important factors were identified
that could explain from 61% to 99% of the observed yield or quality variability in the study site-years. Hybrid was found
to be the most important factor overall for quality in both fields, and for yield as well in Field 1. The relative importance
of soil and landscape factors for corn yield and quality and their relationships differed by hybrid and field. Cation exchange
capacity (CEC) and relative elevation were consistently identified as among the top four most important soil and landscape
factors for both corn yield and quality in both fields in 2000. Aspect and Zn were among the top five most important factors
in Fields 1 and 2, respectively. Compound topographic index (CTI), profile curvature and tangential curvature were, in general,
not important in the study site-years. The response curves generated by the ANN models were more informative than simple correlation
coefficients or coefficients in multiple regression equations. We conclude that hybrid was more important than soil and landscape
factors for consideration in precision crop management, especially when grain quality was a management objective.
While there is a recognised need to adapt to changing climatic conditions, there is an emerging discourse of limits to such
adaptation. Limits are traditionally analysed as a set of immutable thresholds in biological, economic or technological parameters.
This paper contends that limits to adaptation are endogenous to society and hence contingent on ethics, knowledge, attitudes
to risk and culture. We review insights from history, sociology and psychology of risk, economics and political science to
develop four propositions concerning limits to adaptation. First, any limits to adaptation depend on the ultimate goals of
adaptation underpinned by diverse values. Second, adaptation need not be limited by uncertainty around future foresight of
risk. Third, social and individual factors limit adaptation action. Fourth, systematic undervaluation of loss of places and
culture disguises real, experienced but subjective limits to adaptation. We conclude that these issues of values and ethics,
risk, knowledge and culture construct societal limits to adaptation, but that these limits are mutable.
This paper examines the effects of climatic and non-climatic factors on the mean and variance of corn, soybean and winter
wheat yield in southwestern Ontario, Canada over a period of 26years. Average crop yields increase at a decreasing rate with
the quantity of inputs used, and decrease with the area planted to the crop. Climate variables have a major impact on mean
yield with the length of the growing season being the primary determinant across all three crops. Increases in the variability
of temperature and precipitation decrease mean yield and increase its variance. Yield variance is poorly explained by both
seasonal and monthly climate variable models. Projections of future climate change suggest that average crop yield will increase
with warmer temperatures and a longer growing season which is only partially offset by forecast increases in the variability
of temperature and rainfall. The projections would also depend on future technological developments, which have generated
significant increases in yield over time despite changing annual weather conditions.
The increase in atmospheric carbon dioxide concentration and changes in associated climatic variables will likely have a major influence on regional as well as international crop production. This study describes an assessment of simulated potential maize (Zea mays) grain yield using (i) generated weather data and (ii) generated weather data modified by plausible future climate changes under a normal planting date and dates 15 days earlier and 15 days later using CropSyst, a cropping systems simulation model. The analysis is for maize production at Cedara, a summer rainfall location within the midlands of KwaZulu-Natal, South Africa. Baseline weather data input series were generated by a stochastic weather generator, ClimGen, using 30 years of observed weather data (1971–2000). The generated baseline weather data series was similar to the observed for its distributions of daily rainfall and wet and dry day series, monthly total rainfall and its variances, daily and monthly mean and variance of precipitation, minimum and maximum air temperatures, and solar radiant density. In addition, Penman-Monteith daily grass reference evaporation (ETo) calculated using the observed and generated weather data series were similar except that the ETo values between 2 and 3 mm were less for the observed than for the corresponding generated values. Maize grain yields simulated using the observed and generated weather data series with different planting dates were compared. The simulated grain yields for the respective planting dates were not statistically different from each other. However, the grain yields simulated using the generated weather data had a significantly smaller variance than the grain yields simulated using the observed weather data series. The generated baseline weather data were modified by synthesized climate projections to create a number of climatic scenarios. The climate changes corresponded to a doubling of carbon dioxide concentration to 700 μl l−1 without air temperature and water regime changes, and a doubling of carbon dioxide concentration accompanied by mean daily air temperature and precipitation increases of 2 °C and 10%, 2 °C and 20%, 4 °C and 10%, and 4 °C and 20%, respectively. The increase in the daily mean minimum air temperature was taken as three times the increase in daily mean maximum air temperature. Input crop parameters of radiation use and biomass transpiration efficiencies were modified for maize in CropSyst, to account for physiological changes due to increased carbon dioxide concentration. Under increased carbon dioxide concentration regimes, maize grain yields are much more affected by changes in mean air temperature than by precipitation. The results indicate that analysis of the implications of variations in the planting date on maize production may be most useful for site-specific analyses of possible mitigation of the impacts of climate change through alteration of crop management practices.
Agrometeorological information, used for decision making, represents part of a continuum; at the other end is scientific knowledge and understanding. Other components of this continuum are the collection of data and transforming data into useful information. Information has value when it is disseminated in such a way that the end-users get the maximum benefit in applying its content. This paper explores the potential of the new information and communications technologies to improve the access to agrometeorological information. The Internet will play an important role in the collection and transfer of information. In developing countries, Multi-Purpose Community Telecentres (MCTs) will be the equivalent of an information supermarket. Radio can be used to transfer information from MCTs to rural areas. Using response farming as an example, a prototype information system that can have wide applicability is suggested. Procedures on evaluating the impact of agrometeorological information are provided. Future concerns about the information needs of diverse end-users, information on a fee basis, and the training needs of end-users and intermediaries are discussed. Although modern technology has improved agrometeorological information and increased the number of end-users, continued improvements are necessary to ensure that the content of the information is adequate to fulfill the requirements of the farming communities.
Land cover maps increasingly underlie research into socioeconomic and environmental patterns and processes, including global change. It is known that map errors impact our understanding of these phenomena, but quantifying these impacts is difficult because many areas lack adequate reference data. We used a highly accurate, high-resolution map of South African cropland to assess 1) the magnitude of error in several current generation land cover maps, and 2) how these errors propagate in downstream studies. We first quantified pixel-wise errors in the cropland classes of four widely used land cover maps at resolutions ranging from 1 to 100 km, then calculated errors in several representative “downstream” (map-based) analyses, including assessments of vegetative carbon stocks, evapotranspiration, crop production, and household food security. We also evaluated maps’ spatial accuracy based on how precisely they could be used to locate specific landscape features. We found that cropland maps can have substantial biases and poor accuracy at all resolutions (e.g. at 1 km resolution, up to ∼45% underestimates of cropland (bias) and nearly 50% mean absolute error (MAE, describing accuracy); at 100 km, up to 15% underestimates and nearly 20% MAE). National-scale maps derived from higher resolution imagery were most accurate, followed by multi-map fusion products. Constraining mapped values to match survey statistics may be effective at minimizing bias (provided the statistics are accurate). Errors in downstream analyses could be substantially amplified or muted, depending on the values ascribed to cropland-adjacent covers (e.g. with forest as adjacent cover, carbon map error was 200-500% greater than in input cropland maps, but ∼40% less for sparse cover types). The average locational error was 6 km (600%). These findings provide deeper insight into the causes and potential consequences of land cover map error, and suggest several recommendations for land cover map users.
This study assesses changes over the past decade in the farm size distributions of Ghana, Kenya, Tanzania and Zambia. Among all farms below 100 hectares in size, the share of land on small-scale holdings under five hectares has declined except in Kenya. Medium-scale farms (defined here as farm holdings between five and 100 hectares) account for a rising share of total farmland, especially in the 10 to 100 hectare range where the number of these farms is growing especially rapidly. Medium-scale farms control roughly 20% of total farmland in Kenya, 32% in Ghana, 39% in Tanzania, and over 50% in Zambia. The rapid rise of medium-scale holdings in most cases reflects increased interest in land by urban-based professionals or influential rural people. About half of these farmers obtained their land later in life, financed by non-farm income. The rise of medium-scale farms is affecting the region in diverse ways that are difficult to generalize. Many such farms are a source of dynamism, technical change and commercialization of African agriculture. However, medium-scale land acquisitions may exacerbate land scarcity in rural areas, which could have important effects given the projected 60% increase in rural Africa's population between 2015 and 2050. Medium-scale farmers tend to dominate farm lobby groups and influence agricultural policies and public expenditures to agriculture in their favor. Nationally representative Demographic and Health Survey (DHS) data from six countries (Ghana, Kenya, Malawi, Rwanda, Tanzania and Zambia) show that urban households own 5% to 35% of total agricultural land and that this share is rising in all countries where DHS surveys were repeated. This suggests a new and hitherto unrecognized channel by which medium-scale farmers may be altering the strength and location of agricultural growth and employment multipliers between rural and urban areas. Given current trends, medium-scale farms are likely to soon become the dominant scale of farming in many African countries. three anonymous reviewers.
How rainfall arrives, in terms of its frequency, intensity and the timing and duration of rainy season, may have a large influence on rainfed agriculture. However, a thorough assessment of these effects is largely missing. This study combines a new synthetic rainfall model and two independently-validated crop models (APSIM and SARRA-H) to assess sorghum yield response to possible shifts in seasonal rainfall characteristics in West Africa. We find that shifts in total rainfall amount primarily drive the rainfall-related crop yield change, with less relevance to intra-seasonal rainfall features. However, dry regions (total annual rainfall below 500 mm/year) have a high sensitivity to rainfall frequency and intensity, and more intense rainfall events have greater benefits for crop yield than more frequent rainfall. Delayed monsoon onset may negatively impact yields. Our study implies that future changes in seasonal rainfall characteristics should be considered in designing specific crop adaptations in West Africa.
New advances in satellite data acquisition and processing offer promise for monitoring agricultural lands globally. Using these data to estimate crop yields for individual fields would benefit both crop management and scientific research, especially for areas where reliable ground-based estimates are not currently made. Here we introduce a generalized approach for mapping crop yields with satellite data and test its predictions for yields across more than 17,000 maize fields and 11,000 soybean fields spanning multiple states and years in the Midwestern United States. The method, termed SCYM (a scalable satellite-based crop yield mapper), uses crop model simulations to train statistical models for different combinations of possible image acquisition dates, and these are then applied to Landsat and gridded weather data within the Google Earth Engine platform, where the Landsat is composited to find the “best” dates of observations on a pixel-by-pixel basis. SCYM estimates successfully captured a significant fraction of maize yield variation in all state-years, with a range of 14–58% and an average of 35% for this particular study region and crop. Similar results were observed for soybean, with an average of 32% of yield variation captured. The multi-year yield estimates were also used to examine the temporal persistence of yield advantages for the top yielding fields in different counties, which is one measure of how important factors such as farmer skill are in explaining yield gaps. The strength of the SCYM approach lies in its ability to leverage physiological knowledge embedded in crop models to interpret satellite observations in a scalable way, as it can be readily applied to new crops, regions, and types and timing of remote sensing observations without the need for ground calibration.
Short durations of very high spring soil moisture can influence crop yields in many ways, including delaying planting and damaging young crops. The central United States has seen a significant upward trend in the frequency and intensity of extreme precipitation in the 20th century, potentially leading to more frequent occurrences of saturated or nearly saturated fields during the planting season, yet the impacts of these changes on crop yields are not known. Here we investigate the yield response to excess spring moisture for both maize and soybean in the U.S. states of Illinois, Iowa, and Indiana, and the impacts of historical trends for 1950-2011. We find that simple measures of extreme spring soil moisture, derived from fine-scale daily moisture data from the Variable Infiltration Capacity (VIC) hydrologic model, lead to significant improvements in statistical models of yields for both crops. Individual counties experience up to 10 % loss in years with extremely wet springs. However, losses due to historical trends in excess spring moisture measures have generally been small, with 1-3 % yield loss over the 62 year study period.
Livelihood security in Eastern and Southern Africa is strongly dependent on rainfall distribution and land management practices among smallholder farmers. Over 95 % of the land used for food production is based on rainfed agriculture. The major challenge for the rural communities, representing up to 80% of the population in certain countries, is to improve the productivity of the arable land and the available water resources. This paper gives an outline of the hydrological challenges facing smallholder farmers with focus on water scarce areas. The importance of rainfall partitioning rather than rainfall totals is discussed. The main focus is on the management of rural water using low-tech practices, both for domestic purposes and for crop production. Case studies from Eastern and Southern Africa are presented, showing the potential of stabilising the water supply over time both for livestock, household use, and for crop production. The challenges facing research and extension of introducing water management on different scales (household, community, catchment) is discussed.
Increasingly erratic rainfall and unreliable cropping seasons in southern Africa, combined with high food prices, heighten vulnerability of rural people to food insecurity. To understand what actions are needed to expand adaptive capacity to climate change and its consequences for food security, it is useful to learn from existing agricultural practices in semi-arid areas that exploit positive opportunities of rainfall variability. To determine how residents attain food self-sufficiency based on rain-fed maize farming in a semi-arid region that receives an average annual precipitation of 400 mm, we carried out a detailed, interdisciplinary study of the agricultural system in Massingir, Mozambique from 2006 to 2010. We found that some people produced enough maize when rainfall conditions were favorable to sustain the staple food needs of a household for 2–3 years, buffering the negative effects of subsequent poor cropping seasons and avoiding seasonal hunger periods. To maximize production people employed a variety of practices including: planting after every rainfall event throughout the rainy season, up to six times in one season on as large an area as possible, as much as 18 ha per household, and employing labor/oxen exchange arrangements. We explored the role of these practices as key factors that determined total food production and variability among households. Although only 35% of planting events were successful, total seed sown represented only 8.5% of harvest over 15 years. Labor/oxen exchange arrangements allowed disadvantaged households to produce twice as much as without collaboration. Recent invasion of the large grain borer (Prostephanus truncatus), a devastating post-harvest storage insect pest, represents a major new threat to the sustainability of the agricultural system and to food security that could worsen with climate change. Our results suggest that policies aimed at reducing vulnerability to climate change could be improved through a deeper understanding of existing practices.
Over the last decade, governments in eastern and southern Africa have become increasingly involved in grain marketing via strategic reserves and marketing boards. Kenya, Malawi, Zimbabwe, Ethiopia, Tanzania, and Zambia all have one or both of these entities, and their level of involvement in grain marketing has generally increased in recent years. Yet, to date, relatively little is known about how the resurgent activities of strategic grain reserves and marketing boards are affecting market prices. This paper estimates the effects of the Zambia Food Reserve Agency’s (FRA) activities on maize market prices in the country.
Climate change could potentially interrupt progress toward a world without hunger. A robust and coherent global pattern is
discernible of the impacts of climate change on crop productivity that could have consequences for food availability. The
stability of whole food systems may be at risk under climate change because of short-term variability in supply. However,
the potential impact is less clear at regional scales, but it is likely that climate variability and change will exacerbate
food insecurity in areas currently vulnerable to hunger and undernutrition. Likewise, it can be anticipated that food access
and utilization will be affected indirectly via collateral effects on household and individual incomes, and food utilization
could be impaired by loss of access to drinking water and damage to health. The evidence supports the need for considerable
investment in adaptation and mitigation actions toward a “climate-smart food system” that is more resilient to climate change
influences on food security.
Although irrigation in Africa has the potential to boost agricultural productivities by at least 50%, food production on the continent is almost entirely rainfed. The area equipped for irrigation, currently slightly more than 13 million hectares, makes up just 6% of the total cultivated area. More than 70% of Africa’s poor live in rural areas and mostly depend on agriculture for their livelihoods. As a result, agricultural development is key to ending poverty on the continent. Many development organizations have recently proposed to significantly increase investments in irrigation in the region. However, the potential for irrigation investments in Africa is highly dependent upon geographic, hydrologic, agronomic, and economic factors that need to be taken into account when assessing the long-term viability and sustainability of planned projects. This paper analyzes the large, dam-based and small-scale irrigation investment potential in Africa based on agronomic, hydrologic, and economic factors. We find significant profitable irrigation potential for both small-scale and large-scale systems. This type of regional analysis can guide distribution of investment funds across countries and should be a first step prior to in-depth country- and local-level assessment of irrigation potential, which will be important to agricultural and economic development in Africa.
The accuracy of supervised classification is dependent to a large extent on the training data used. The aim is often to capture a large training set to fully describe the classes spectrally, commonly with the requirements of a conventional statistical classifier in-mind. However, it is not always necessary to provide a complete description of the classes, especially if using a support vector machine (SVM) as the classifier. A SVM seeks to fit an optimal hyperplane between the classes and uses only some of the training samples that lie at the edge of the class distributions in feature space (support vectors). This should allow the definition of the most informative training samples prior to the analysis. An approach to identify informative training samples was demonstrated for the classification of agricultural classes in south-western part of Punjab state, India. A small, intelligently selected, training data set was acquired in the field with the aid of ancillary information. This data set contained the data from training sites that were predicted before the classification to be amongst the most informative for a SVM classification. The intelligent training collection scheme yielded a classification of comparable accuracy, ~91%, to one derived using a larger training set acquired by a conventional approach. Moreover, from inspection of the training sets it was apparent that the intelligently defined training set contained a greater proportion of support vectors (0.70), useful training sites, than that acquired by the conventional approach (0.41). By focusing on the most informative training samples, the intelligent scheme required less investment in training than the conventional approach and its adoption would have reduced total financial outlay in classification production and evaluation by ~26%. Additionally, the analysis highlighted the possibility to further reduce the training set size without any significant negative impact on classification accuracy.
Remote sensing allows scientists to detect slowly evolving natural hazards such as agricultural drought. Famine early warning systems transform these data into actionable policy information, enabling humanitarian organizations to respond in a timely and appropriate manner.
These life‐saving responses are increasingly important: In 2006, one out of eight people did not have enough food to eat and 22 million more people became sufficiently undernourished to require intervention, prompting 22 countries to provide $6.5 billion in food aid. Since their inception in the mid‐1980s, the combination of monitoring and mitigation systems has dramatically reduced the number of famines caused by biophysical hazards, such as floods, drought, and pests, that destroy food crops [Murphy and McAfee, 2005]. Yet despite this notable achievement, many countries, mostly in Africa, face chronic and increasing food insecurity.
Given the accumulating evidence of climate change in sub-Saharan Africa, there is an urgent need to develop more climate resilient maize systems. Adaptation strategies to climate change in maize systems in sub-Saharan Africa are likely to include improved germplasm with tolerance to drought and heat stress and improved management practices. Adapting maize systems to future climates requires the ability to accurately predict future climate scenarios in order to determine agricultural re-sponses to climate change and set priorities for adaptation strategies. Here we review the projected climate change scenarios for Africa's maize growing regions using the out-puts of 19 global climate models. By 2050, air temperatures are expected to increase throughout maize mega-environ-ments within sub-Saharan Africa by an average of 2.1°C. Rainfall changes during the maize growing season varied with location. Given the time lag between the development of improved cultivars until the seed is in the hands of farmers and adoption of new management practices, there is an urgent need to prioritise research strategies on climate change resilient germplasm development to offset the pre-dicted yield declines.
Southern Africa is a region facing multiple stressors, including chronic, recurrent food insecurity and persistent threats of famine. Climate information, including seasonal climate forecasts, has been heralded as a promising tool for early-warning systems and agricultural risk management in southern Africa. Nevertheless, there is concern that climate information, for example climate forecasts, are not realizing their potential value in the region. The present study considers the actual and potential roles played by climate information in reducing food insecurity in southern Africa from 2 perspectives. The first relates to improved understanding of the contextual environment in which end users operate and use information. Users, including farmers, usually operate in an environment of considerable uncertainty, reacting to and coping with multiple stressors whose impacts are not always clear or predictable. The second perspective relates to improving the current design and variety of mechanisms (e.g. climate outlook forums) for the dissemination and uptake of climate information. The first relates to improved understanding of the contextual environment in which end users operate and use information. Users, including farmers, usually operate in an environment of considerable uncertainty, reacting to and coping with multiple stressors whose impacts are not always clear or predictable. The second perspective relates to improving the current design and variety of mechanisms (e.g. climate outlook forums) for the dissemination and uptake of climate information. Climate information, it is argued, used in isolation (e.g. in 'stand alone' climate outlook forums) and undertaken in a traditional, linear fashion, where information is moved from producer to user, is divorced from the broader, complex social context in which such information is embedded. This current articulation of climate information flow represents an ineffective means of dealing with climate variability and food security. Alternative modes of interaction (e.g. using existing platforms to 'piggyback' information or seeking appropriate 'boundary organisations') should be found to sustainably manage climate risks in the region.
Securing sustainable livelihood conditions and reducing the risk of outmigration in savanna ecosystems hosted in the tropical semiarid regions is of fundamental importance for the future of humanity in general. Although precipitation in tropical drylands, or savannas, is generally more significant than one might expect, these regions are subject to considerable rainfall variability which causes frequent periods of water deficiency. This paper addresses the twin problems of “drought and desertification” from a water perspective, focusing on the soil moisture (green water) and plant water uptake deficiencies. It makes a clear distinction between long-term climate change, meteorological drought, and agricultural droughts and dry spells caused by rainfall variability and land degradation. It then formulates recommendations to better cope with and to build resilience to droughts and dry spells. Coping with desertification requires a new conceptual framework based on green-blue water resources to identify hydrological opportunities in a sea of constraints. This paper proposes an integrated land/water approach to desertification where ecosystem management supports agricultural development to build social-ecological resilience to droughts and dry spells. This approach is based on the premise that to combat desertification, focus should shift from reducing trends of land degradation in agricultural systems to water resource management in savannas and to landscape-wide ecosystem management.
Drought is the most critical environmental factor limiting the productivity of agricultural crops worldwide. Increased frequency and severity of drought are expected to accompany climate change and will negatively impact global food security. Wide yield variability from field to field, and consequently reduced average yield on a regional scale often occur under drought conditions. The reasons for the yield variability are still poorly understood. In this study, we explored sources of soybean yield variability among fields in a rural village of Northeast China associated with a severe drought growing season in 2007. Soil parameter measurements were made on fields following three transects with different distances from homestead. Management data were assembled from household interviews. The relative importance of soil parameters and management practices resulting in yield variability among fields was analyzed with general linear model (GLM) and classification and regression trees (CARTs) models. Our analysis showed that variability in management options, as opposed to variability in soil parameters, caused the majority of yield variability from field to field. The amount of P applied was the most important variable determining yield variability and explained roughly 61% of the variability. Whether or not manure was added into fields was of secondary importance. The classification tree analysis indicated that yield differences among transects was attributed to the content of K nutrient. This might result from variations of long-term management options with distance from homestead. CART models are robust technique for predicting yield variability responses to variations of soil properties and management practices due to its low prediction error. Our study highlights the pressing need to adjust management strategies for narrowing yield variability and increasing crop production in drought years. We recommend that in addition to testing soil, government programs in China should also pay close attention to management practices of farmers.
This paper begins with an overview of the African rainfall regime, noting in particular the contrast among various regions of the continent, followed by a description of the nature of climatic (i.e., rainfall) variability over Africa on time scales of decades and centuries. The decadal scale is examined using modern data covering the twentieth century. The century scale is examined using historical reconstructions of climate, based on a combination of geologic, geographic and historical information (e.g., lake chronologies, landscape descriptions, archives and diaries).The presentation includes some results of an analysis of a new historical semi-quantitative data set for Africa covering the last two centuries. It was produced using a combination of historical information, nineteenth century rainfall records, and statistical relationships among various sectors of Africa. Presented here are reconstructions of lake level fluctuations for numerous lakes of eastern and southern Africa.This overview of climatic fluctuations is utilized to uncover inherent spatial and temporal characteristics of the rainfall variability. The dominance over time of various spatial modes is emphasized and the questions of synchroneity of the hemispheres and the abruptness of change are considered. The contrast between the two hemispheres is also surveyed, notably the different time scales of variability and potential causal factors in the variability. One of the most important contrasts is the multi-decadal persistence of anomalies over most of northern Africa. This has implications for the causes of long-term fluctuations, even those historical and paleo-time scales.
To guide soil fertility investment programmes in sub-Saharan Africa, better understanding is needed of the relative importance of soil and crop management factors in determining smallholder crop yields and yield variability. Spatial variability in crop yields within farms is strongly influenced by variation in both current crop management (e.g. planting dates, fertilizer rates) and soil fertility. Variability in soil fertility is in turn strongly influenced by farmers’ past soil and crop management. The aim of this study was to investigate the relative importance of soil fertility and crop management factors in determining yield variability and the gap between farmers’ maize yields and potential yields in western Kenya. Soil fertility status was assessed on 522 farmers’ fields on 60 farms and paired with data on maize-yield and agronomic management for a sub-sample 159 fields. Soil samples were analysed by wet chemistry methods (1/3 of the samples) and also by near infrared diffuse reflectance spectroscopy (all samples). Spectral prediction models for different soil indicators were developed to estimate soil properties for the 2/3 of the samples not analysed by wet chemistry. Because of the complexity of the data set, classification and regression trees (CART) were used to relate crop yields to soil and management factors. Maize grain yields for fields of different soil fertility status as classified by farmers were: poor, 0.5–1.1; medium, 1.0–1.8; high, 1.4–2.5 t ha−1. The CART analysis showed resource use intensity, planting date, and time of planting were the principal variables determining yield, but at low resource intensity, total soil N and soil Olsen P became important yield-determining factors. Only a small group of plots with high average grain yields (2.5 t ha−1; n = 8) was associated with use of nutrient inputs and good plant stands, whereas the largest group with low average yields (1.2 t ha−1; n = 90) was associated with soil Olsen P values of less than 4 mg kg−1. This classification could be useful as a basis for targeting agronomic advice and inputs to farmers. The results suggest that soil fertility variability patterns on smallholder farms are reinforced by farmers investing more resources on already fertile fields than on infertile fields. CART proved a useful tool for simplifying analysis and providing robust models linking yield to heterogeneous crop management and soil variables.
Policies to promote adaptation climate risks often rely on the willing cooperation of the intended beneficiaries. If these beneficiaries disagree with policy makers and program managers about the need for adaptation, or the effectiveness of the measures they are being asked to undertake, then implementation of the policies will fail. A case study of a resettlement program in Mozambique shows this to be the case. Farmers and policy makers disagreed about the seriousness of climate risks, and the potential negative consequences of proposed adaptive measures. A project to provide more information about climate change to farmers did not change their beliefs. The results highlight the need for active dialog across stakeholder groups, as a necessary condition for formulating policies that can then be successfully implemented.
Understanding the variability of the terrestrial hydrologic cycle is central to determining the potential for extreme events and susceptibility to future change. In the absence of long-term, large-scale observations of the components of the hydrologic cycle, modeling can provide consistent fields of land surface fluxes and states. This paper describes the creation of a global, 50-yr, 3-hourly, 1.0° dataset of meteorological forcings that can be used to drive models of land surface hydrology. The dataset is constructed by combining a suite of global observation-based datasets with the National Centers for Environmental PredictionâNational Center for Atmospheric Research (NCEPâNCAR) reanalysis. Known biases in the reanalysis precipitation and near-surface meteorology have been shown to exert an erroneous effect on modeled land surface water and energy budgets and are thus corrected using observation-based datasets of precipitation, air temperature, and radiation. Corrections are also made to the rain day statistics of the reanalysis precipitation, which have been found to exhibit a spurious wavelike pattern in high-latitude wintertime. Wind-induced undercatch of solid precipitation is removed using the results from the World Meteorological Organization (WMO) Solid Precipitation Measurement Intercomparison. Precipitation is disaggregated in space to 1.0° by statistical downscaling using relationships developed with the Global Precipitation Climatology Project (GPCP) daily product. Disaggregation in time from daily to 3 hourly is accomplished similarly, using the Tropical Rainfall Measuring Mission (TRMM) 3-hourly real-time dataset. Other meteorological variables (downward short- and longwave radiation, specific humidity, surface air pressure, and wind speed) are downscaled in space while accounting for changes in elevation. The dataset is evaluated against the bias-corrected forcing dataset of the second Global Soil Wetness Project (GSWP2). The final product provides a long-term, globally consistent dataset of near-surface meteorological variables that can be used to drive models of the terrestrial hydrologic and ecological processes for the study of seasonal and interannual variability and for the evaluation of coupled models and other land surface prediction schemes.