Machine-learning approach to crop yield prediction with the spatial extent of drought

Crop yield is one of the variables used to assess the impact of droughts on agriculture. Crop growth models calculate yield and variables related to plant development and become more suitable for crop yield estimation. However, these models are limited in that specific data are needed for computation. Given this limitation, machine learning (ML) models are often widely utilised instead, but their use with the spatial characteristics of droughts as input data is limited. This research explored the spatial extent of drought (area) as input data for building an approach to predict seasonal crop yield. This ML approach is made up of two components. The first includes polynomial regression (PR) models, and the second considers artificial neural network (ANN) models. In this approach, the purpose was to evaluate both types of ML models (PR and ANN) and integrate them into one operational tool. The logic is as follows: ANN models determine the most accurate predictions, but in practice, issues regarding data retrieval and processing can make the use of equations, i.e. PR, preferable. The proposed approach provides these PR equations to perform such calculations with early and preliminary input. The estimates can be further improved when the ANN models are run with the final input data. The results indicated that the empirical equations (PR) produced good predictions when using drought area as the input. ANN provides better estimates, in general. This research will improve drought monitoring systems for assessing drought effects. Since it is currently possible to calculate drought areas within these systems, the direct application of the prediction of drought effects is possible to integrate by following approaches such as the one presented or similar.

37 Years of Forest Monitoring in Switzerland: Drought Effects on Fagus sylvatica

European beech is one of the most important deciduous tree species in natural forest ecosystems in Central Europe. Its dominance is now being questioned by the emerging drought damages due to the increased incidence of severe summer droughts. In Switzerland, Fagus sylvatica have been observed in the Intercantonal Forest Observation Program since 1984. The dataset presented here includes 179176 annual observations of beech trees on 102 plots during 37 years. The plots cover gradients in drought, nitrogen deposition, ozone, age, altitude, and soil chemistry. In dry regions of Switzerland, the dry and hot summer of 2018 caused a serious branch dieback, increased mortality in Fagus sylvatica and increased yellowing of leaves. Beech trees recovered less after 2018 than after the dry summer 2003 which had been similar in drought intensity except that the drought in 2018 started earlier in spring. Our data analyses suggest the importance of drought in subsequent years for crown transparency and mortality in beech. The drought in 2018 followed previous dry years of 2015 and 2017 which pre-weakened the trees. Our long-term data indicate that the drought from up to three previous years were significant predictors for both tree mortality and for the proportion of trees with serious (>60%) crown transparency. The delay in mortality after the weakening event suggests also the importance of weakness parasites. The staining of active vessels with safranine revealed that the cavitation caused by the low tree water potentials in 2018 persisted at least partially in 2019. Thus, the ability of the branches to conduct water was reduced and the branches dried out. Furthermore, photooxidation in light-exposed leaves has increased strongly since 2011. This phenomenon was related to low concentrations of foliar phosphorus (P) and hot temperatures before leaf harvest. The observed drought effects can be categorized as (i) hydraulic failure (branch dieback), (ii) energy starvation as a consequence of closed stomata and P deficiency (photooxidation) and (iii) infestation with weakness parasites (beech bark disease and root rots).

Assessing Drought Effects on Banana Production and On-Farm Coping Strategies by Farmers – A Study in the Cattle Corridor of Uganda

Drought is a major threat to banana production in Uganda, leading to large yield losses. This study documented drought effects on banana production and identified farmers’ coping strategies to mitigate the impact of droughts. Interviews were conducted in eight districts, randomly selected from banana-growing districts in Uganda’s cattle corridor, characterised by frequent droughts. Data were collected from 120 respondents/farms. Banana production in the study area was dominated by small-scale farmers, growing mostly a combination of cooking and dessert banana types. Among the 15 identified effects of drought stress on banana growth, reduced bunch weight, wilting and drying of leaves, reduced leaf production and reduced number of fingers and clusters were the most reported. ‘Mpologoma’ and ‘FHIA 17’ cultivars were reported as the most and least affected by drought stress, respectively. Although the cattle corridor is prone to recurrent droughts, the deployment of drought coping strategies was mostly low, with farmers using one to three strategies. A total of 12 drought mitigation practices were used across the cattle corridor, with mulching being the most common option. Irrigation was perceived as the most effective mitigation option though its deployment was limited by water scarcity and the high cost of water pumps. This study suggests the need for government support in drought stress mitigation practices and the development of drought-tolerant cultivars by breeders. Additionally, farmers need to prioritise preventive coping strategies like planting drought-tolerant cultivars, irrigation, mulching, and manure application and ensure the appropriate time of deployment of mitigation practices.