Crop Calendar Research Articles

Strengthening adaptive capacities of small holding South Asian agrarian community through Climate Information Network -based decision support tool

Marginal farmers in rain-fed agrarian systems can benefit immensely from real-time contingency planning and decisionmaking tools in the climate milieu. Long-term climate anomaly prediction of some selected locations was made. Forty-two Global Circulation Models (GCM) from the Coupled Model Intercomparison Project Phase 5 project (CMIP5) were used for the present work and Representative Concentration Pathway (RCP) 4.5 to prepare future projections. The effect of the climate anomaly on existing farming practices in two selected climate-vulnerable locations of Purulia, India, and Shyamnagar, Bangladesh, has been analysed. A participatory action research and Climate Information Network (CIN)-based crop calendar has been devised for the selected locations. An interactive Android bot in the form of a Decision-Making Cropping Support System (DMCSS) has been devised and field tested. This system can provide seasonal crop suggestions and suggest contingency management to farmers. The system also provides risk assessments on the basis of preparedness and categorical inputs that users provide on crop, climate and contrivances.

Cropland Mapping in Tropical Smallholder Systems with Seasonally Stratified Sentinel-1 and Sentinel-2 Spectral and Textural Features

Mapping arable field areas is crucial for assessing agricultural productivity but poses challenges in sub-Saharan agroecosystems because of diverse crop calendars, small and irregularly shaped fields, persistent cloud cover, and lack of high-quality model training data. This study proposes several methodological improvements to overcome these challenges. Specifically, it utilizes long-term MODIS data to stratify finer Sentinel-2 reflectance and Sentinel-1 backscatter image features on a per-pixel basis. It also incorporates texture features and employs a machine learning approach with over 300,000 samples. The eastern region of Ghana was stratified into seven seasonal strata exhibiting distinct vegetation seasonality, capturing diversity in crop calendars, using long-term MODIS (2001–2009) normalized difference vegetation index phenology. Three years (2017–2019) of Sentinel-1 and Sentinel-2 original bands at 20 m were composited into dry and wet seasonal features according to the strata, from which spectral, polarimetric, and texture features were extracted. The field boundaries were digitized using PlanetScope images (2018–2019). Random Forest classifier with 10-fold cross-validation and recursive feature elimination was used for feature selection and model building. Including topographic variables, out of 137 image features, only 11 features were found important. Sentinel-2 SWIR-based spectral features were most important, followed by Sentinel-1 polarimetric (VV) and elevation features. Half of the 11 features were variance texture features, followed by spectral features. The Random Forest classifier produced a 0.78 AUC score with overall precision, recall, and F1-score of 0.96, 0.78, and 0.85, respectively. While the precision for both classes was >0.90, the recall rate for arable areas was half that of non-arable areas. Future studies could improve the technical workflow with reliable balanced sampling, narrowband hyperspectral images, and fully polarized SAR images.

GIS Based Mapping for a Better Crop Planning for Perumpadappu Block of Ponnani Kole Lands in Kerala

Kole lands are one of the biggest saline, humid tropical wetland ecosystems. It spreads across Thrissur and Malappuram districts of Kerala state. They serve as a water storage structure during rainy season. After rains, the stored water is dewatered and circulated in various Kole padasekarams and channels to cultivate paddy. However, the problem is the lack of a proper water management plan resulting in crop loss. The current study is conducted in the Perumpadappu block of Ponnani Kole lands in Malappuram district. In this study, both GIS and Remote sensing technology have been collaborated to plan the crop calendar. The cropping calendar of Ponnai Kole was developed mainly for 37 Kole padashekharams of Perumpadappu block. Data about elevation, transportation, irrigation, weather and soil were collected to arrive at the crop calendar. Together, by adopting participatory rural appraisal methods and local knowledge were integrated. The final crop calendar was wetted with farmers and experts.

Assessing farmers’ perception based composite drought vulnerability in Godavari Middle Sub-basin, India

Climate change has increased the frequency and intensity of drought in various parts of the world. Godavari Middle Sub-basin of India has been experiencing significant change in rainfall, maximum and minimum temperature. Farmers of Sub-basin depend on mercy of rainfall for their cultivation. Occurrence of drought lead to farmers suicide due to crop failure and losses of agricultural production. Therefore, composite drought vulnerability assessment is essential in the Sub-basin. This paper makes an initial attempt to quantify composite vulnerability to drought using socio-economic and environmental indicators at household level. Stratified random sampling was utilized to collect primary data. A total number of 450 households were surveyed. The composite vulnerability index calculation was based on 5 domains namely economic, social, environmental, ecological, adaption and mitigation strategies. Results revealed that highest drought vulnerability was found in marginal farmers among all the types of farmers due to high climate change vulnerability and low adaptation strategies. Low level of education, low annual income and high dependency on local money lenders were the main causes of high drought vulnerability for marginal farmers. High adaptation strategies helped large farmer to mitigate the impact of drought. The adaptations and mitigation strategies encompassed change in crops, diversification of crops, keeping land unsown after the possibility of drought, change in traditional irrigation practices to sprinkler and drip irrigation, change in crop calendar and sowing of drought-tolerant crops. Community-based preparedness planning could be effective to mitigate the impact of climate change induced drought vulnerability in Godavari Middle Sub-basin.

Improving crop yield prediction accuracy by embedding phenological heterogeneity into model parameter sets

CONTEXTThe assimilation of Remote Sensing (RS) data into crop models improves the accuracy of yield predictions by considering crop growth dynamics and their spatial heterogeneity due to the different management practices and environmental conditions. OBJECTIVEThis study proposes a new method for performing sub-regional yield predictions (Nomenclature of territorial units for statistics, NUTS-3 level) using RS time series data and crop models. The primary objective was to release a procedure whereby the heterogeneity of the agricultural landscape (i.e. the agrophenotypes) observed from RS is used to drive crop model calibration. METHODSFine-resolution barley and maize distribution maps (100 m) and related crop calendars have been used to derive the “where” and “when” of the 8-day MODIS NDVI time series data, located in Apulia, Tuscany, and Veneto in 2018–2019. Principal Component Analysis and Hierarchical Clustering have been applied to NDVI seasonal profiles to identify the agrophenotypes, which have been used to derive crop growth and leaf area index dynamics. These data served as the reference to optimize the most relevant parameters of the WOFOST_GT model, using gridded weather data as input (0.25° resolution, Copernicus ERA5). Parameter distributions from multiple automatic calibrations have been sampled to characterize the agricultural heterogeneity within 22 NUTS-3 administrative units. Yield statistics from the Italian National Institute of Statistics have been used as reference data to test the accuracy of the yield simulation by the crop model WOFOST_GT. RESULTS AND CONCLUSIONSThe agrophenotypes reflected the wide north-south latitudinal gradient experienced in Italy by the two crops, leading to a gap of 15–30 days in barley and maize flowering and harvest dates across the study area. Average Nast-Sutcliffe modelling efficiency in reproducing LAI dynamics from RS (0.6) and Relative Root Mean Square Error (RRMSE) in predicting yield data (12.1% for barley, 3.7% for maize) in hindcast simulations demonstrated the effectiveness of our approach. The average RRMSE was reduced by 32.7% (barley) and 8.5% (maize) compared to baseline, decreasing by 0.7–1 Mg ha−1 absolute yield errors on both crops. SIGNIFICANCEThe inclusion of local agrophenotypes in the yield prediction workflow reduced errors in yield prediction compared to unsupervised simulations at NUTS-3 level. The script for agrophenotypes extraction and the model parameter sets are released to the scientific community, to foster improvements and further applications to other crops, ecoclimatic regions, satellite sensors and spatial scales.

Climate-Adaptive Potential Crops Selection in Vulnerable Agricultural Lands Adjacent to the Jamuna River Basin of Bangladesh Using Remote Sensing and a Fuzzy Expert System

Agricultural crop production was affected worldwide due to the variability of weather causing floods or droughts. In climate change impacts, flood becomes the most devastating in deltaic regions due to the inundation of crops within a short period of time. Therefore, the aim of this study was to propose climate-adaptive crops that are suitable for the flood inundation in risk-prone areas of Bangladesh. The research area included two districts adjacent to the Jamuna River in Bangladesh, covering an area of 5489 km2, and these districts were classified as highly to moderately vulnerable due to inundation by flood water during the seasonal monsoon time. In this study, first, an inundation vulnerability map was prepared from the multicriteria analysis by applying a fuzzy expert system in the GIS environment using satellite remote sensing datasets. Among the analyzed area, 42.3% was found to be highly to moderately vulnerable, 42.1% was marginally vulnerable and 15.6% was not vulnerable to inundation. Second, the most vulnerable areas for flooding were identified from the previous major flood events and cropping practices based on the crop calendar. Based on the crop adaptation suitability analysis, two cash crops, sugarcane and jute, were recommended for cultivation during major flooding durations. Finally, a land suitability analysis was conducted through multicriteria analysis applying a fuzzy expert system. According to our analysis, 28.6% of the land was highly suitable, 27.9% was moderately suitable, 19.7% was marginally suitable and 23.6% of the land was not suitable for sugarcane and jute cultivation in the vulnerable areas. The inundation vulnerability and suitability analysis proposed two crops, sugarcane and jute, as potential candidates for climate-adaptive selection in risk-prone areas.

LAND USE/LAND COVER CLASSIFICATION IN A HETEROGENEOUS AGRICULTURAL LANDSCAPE USING PLANETSCOPE DATA

Abstract. Land use and land cover (LULC) classification has long been an essential topic in Earth Observation research and plays a key role in the sustainable development of agriculture. This study evaluated the accuracy of LULC classification based on an initial clustering step in a heterogeneous agricultural landscape using PlanetScope imagery while checking for variability among their Normalized Difference Vegetation Index (NDVI) temporal signatures. We adopt an object-based image analysis to generate image-objects and then extract statistical information of PlanetScope spectral bands and vegetation indices as input information for classification. The exploratory analysis focused on the double crop class and calculated the distance between NDVI temporal signatures of paired land parcels. We applied an unsupervised clustering technique along with Random Forest algorithm based on multiple tests to classify and analyse gains and losses in accuracies produced by these approaches. Our results showed that the initial clustering method outperformed the non-clustered classification of LULC in overall accuracy measures. The exploratory analysis demonstrated that double crops might present high intra-class variability and diverse crop calendars for neighbour land parcels. The accuracies achieved represent promising opportunities for the sufficiently accurate classification of such areas, and the knowledge of the intra-class variability allows the analyst to infer the temporal dynamics of crop fields. We reinforce that further work could assess other types of classifiers, especially in areas with a large number of crop types and distinct management practices.

High-Resolution Mapping of Seasonal Crop Pattern Using Sentinel Imagery in Mountainous Region of Nepal: A Semi-Automatic Approach

Sustainable agricultural management requires knowledge of where and when crops are grown, what they are, and for how long. However, such information is not yet available in Nepal. Remote sensing coupled with farmers’ knowledge offers a solution to fill this gap. In this study, we created a high-resolution (10 m) seasonal crop map and cropping pattern in a mountainous area of Nepal through a semi-automatic workflow using Sentinel-2 A/B time-series images coupled with farmer knowledge. We identified agricultural areas through iterative self-organizing data clustering of Sentinel imagery and topographic information using a digital elevation model automatically. This agricultural area was analyzed to develop crop calendars and to track seasonal crop dynamics using rule-based methods. Finally, we computed a pixel-level crop-intensity map. In the end our results were compared to ground-truth data collected in the field and published crop calendars, with an overall accuracy of 88% and kappa coefficient of 0.83. We found variations in crop intensity and seasonal crop extension across the study area, with higher intensity in plain areas with irrigation facilities and longer fallow cycles in dry and hilly regions. The semi-automatic workflow was successfully implemented in the heterogeneous topography and is applicable to the diverse topography of the entire country, providing crucial information for mapping and monitoring crops that is very useful for the formulation of strategic agricultural plans and food security in Nepal.

TRANCO: Thermo radiometric normalization of crop observations

Crop type maps are essential for a wide range of applications such as crop monitoring, and yield estimation. In addition, Earth Observation (EO) systems allow robust and timely mapping of the earth’s surface, usually based on time-series. Yet, existing crop type maps are either global at coarse spatial resolution, or have a local or regional scope. The reasons for this gap can be linked to the scarcity of global crop type datasets at field level to train the models, their bias towards the Northern Hemisphere, as well as the limited transferability of existing crop type models across different regions. One of the main limitations on the transferability is driven by the phenological shift of the crops’ radiometric time series detected by Earth Observation (EO) systems, which is mainly induced by the different climates across regions. In this study, we explore the normalization of EO-based wheat time series with the accumulation of Growing Degree Days (GDD): the Thermo-RAdiometric Normalization of Crop Observations (TRANCO) system. The TRANCO system is based on the assumption that crop phenology evolution is mainly driven by temperature accumulation, represented by the accumulated GDD from Start of Season (SOS) to End of Season (EOS) dates, derived from a crop calendar. We tested the proposed method to normalize wheat on a database of globally distributed samples, whose results show a great improvement of the GDD F1 score (0.90) compared to a simpler normalization approach based on Time windows defined from SOS calendars (0.87) and a baseline without normalization (0.83).

Indigenous Knowledge of Peanut Cultivation of the Ethnic Khmer Resident in the Mountainous Area in An Giang Province, Vietnam

Indigenous knowledge (IK) can provide useful information on local people's activities within environmental contexts. For countries vulnerable to climate change as Vietnam, research on IK might assist in revealing information about local people's responses to environmental stresses and potentially support decision-making. To enrich the study context in this field, this study aims at determining the IK of peanut cultivation of the Khmer ethnic groups. The study site is in the mountainous area in An Giang province, Vietnam, where the Khmer people are located. The methods of documentary collection, focus group discussion, semi-structured interviews, and experiments were applied in this study. The investigation results showed the innovation of local people in the cultivation. For instance, applying cow manure to a sandy soil can reduce chemical fertilizers and improve peanut yield. The IK reported that the application of the broadcasting method could achieve the cultivation economic profit approximately 2.5 times higher than the dibbling method. The former method is mostly preferred due to its time/labor saving and the growing aged population in the study location. The experimental results indicated that the combination of indigenous and scientific knowledge in peanut cultivation, such as chemical fertilizer application, crop calendar, crop rotation, and varieties, could bring high economic efficiency, improve soil nutrients, and ability to adapt to climate change. The research approaches in this study comprehensively revealed the IK, which can be potentially applicable to similar studies.

Adaptive capacity of high- and low dyke farmers to hydrological changes in the Vietnamese Mekong delta

This research assesses the adaptive capacity of farmers in the Vietnamese Mekong Delta's floodplains (VMD) with respect to hydrological changes. Currently, climate change and socio-economic developments induce extreme- and diminishing floods, which in turn increase farmers' vulnerability. This research assesses farmers' adaptive capacity to hydrological changes using two prevalent farming systems: high dykes featuring triple-crop rice farming and low dykes where fields are left fallow during the flood season. We examine (1) farmers' perceptions on a changing flood regime and their current vulnerabilities and (2) farmers' adaptive capacity through five sustainability capitals. Methods include a literature review and qualitative interviews with farmers. Results show that extreme floods are becoming less frequent and damaging, depending on arrival time, depth, residence time, and flow velocity. In extreme floods, farmers' adaptive capacity is generally strong, and only low dyke farmers experience damage. As for diminishing floods, which is an emerging phenomenon, the overall adaptive capacity of farmers is remarkably weaker and varies between high- and low dyke farmers. Financial capital is lower for low dyke farmers due to their double-crop rice system, and natural capital is low for both farmer groups due to a decrease in soil- and water quality, affecting yields and increasing investment costs. Farmers also struggle with an unstable rice market due to strong fluctuating prices for seeds, fertilizers, and other inputs. We conclude that both high- and low dyke farmers have to cope with new challenges, including fluctuating flood patterns and the depletion of natural resources. Increasing farmers resilience should focus on exploring better crop varieties, adjusting crop calendars, and shifting to less water-intensive crops.

Evaluation of the accuracy of satellite-based rainfed wheat yield dataset over an area with complex geography

Global gridded datasets have a potential of being a reliable resource for filling gaps in observed data in global and regional studies. Global gridded cropping system datasets are useful in various fields, such as crop and hydrological simulations. They could serve as a source alternative to observed datasets. This study evaluates data on rainfed wheat yield derived from a satellite-based Global Dataset of Historical Yields (GDHY) over Iran from the years 1983–2016 (in this period, the average rainfed wheat yield over Iran was 0.86 tons/ha). The process of development of this dataset employed satellite products, solar radiation, harvested area, crop calendar, and production share by cropping season. Our evaluations were conducted at the country scale, province scale, and in reference to different climates based on the Köppen-Geiger climate classification. The index of agreement over Iran is approximately 0.79, and according to percentage bias, GDHY has a 1.6% error over Iran. The dataset was reliable at the sub-national level, over the western parts, and particularly in the mountainous region. In reference to the root mean square error (tons/ha), it showed the lowest performance in the humid part of Iran (Mazandaran, Gilan, and Golestan provinces). Drought analysis showed that the reference crop yield dataset over the study area is more consistent with SPI-12, although GDHY agrees with SPI-3. In contrast to drier years in recent years, the crop yield has increased. This positive trend can be justified by increasing the rate of fertilizer applications, better grazing, an overall improvement in farming technologies, and better farm management at the country scale. In conclusion, GDHY can be used at the national scale for different analyses or modeling, but it is inaccurate at the sub-national level.

Performance evaluation of buried PVC pipelines for tank irrigation: a case study of Deevanur tank

Abstract Water is vital to life and development in all parts of the world. Since rain or other precipitation is the only source of water, it should be carefully harvested, stored, conserved and managed for the benefit of the largest number by optimizing the use of the amount available. The available water should be utilized judiciously so that the future generation is saved from deterioration stress. This study was conducted in the Deevanur tank located in Tindivanam Taluk in Tamil Nadu, India. The study was carried out in the crop calendar years from 2006–07 to 2017–18. The performance of the tank irrigation system in this study was evaluated using a Cropping Intensity indicator. This study reveals that Cropping Intensity was increased to more than 100% after rehabilitation of the tank irrigation system and is easily observed. Before rehabilitation of the tank, the average Cropping Intensity value was only 0.87. After rehabilitation of the tank, the average Cropping Intensity value increased to 1.20. The Cropping Intensity indicator is good for evaluating the modernization of the tank irrigation system.

Optimality-based modelling of wheat sowing dates globally

CONTEXTSowing dates are currently an essential input for crop models. However, in the future, the optimal sowing time will be affected by climate changes and human adaptations to these changes. A better understanding of what determines the choice of wheat type and sowing dates is required to be able to predict future crop yields reliably. OBJECTIVEThis study was conducted to understand how climate conditions affect the choice of wheat types and sowing dates globally. METHODSWe develop a model integrating optimality concepts for simulating gross primary production (GPP) with climate constraints on wheat phenology to predict sowing dates. We assume that wheat could be sown at any time with suitable climate conditions and farmers would select a sowing date that maximises yields. The model is run starting on every possible climatically suitable day, determined by climate constraints associated with low temperature and intense precipitation. The optimal sowing date is the day which gives the highest yield in each location. We evaluate the simulated optimal sowing dates with data on observed sowing dates created by merging census-based datasets and local agronomic information, then predict their changes under future climate scenarios to gain insight into the impacts of climate change. RESULTS AND CONCLUSIONSCold-season temperatures are the major determinant of sowing dates in the extra-tropics, whereas the seasonal cycle of monsoon rainfall is important in the tropics. Our model captures the timing of reported sowing dates, with differences of less than one month over much of the world; maximum errors of up to two months occur in tropical regions with large altitudinal gradients. Discrepancies between predictions and observations are larger in tropical regions than temperate and cold regions. Slight warming is shown to promote earlier sowing in wet areas but later in dry areas; larger warming leads to delayed sowing in most regions. These predictions arise due to the interactions of several influences on yield, including the effects of warming on growing-season length, the need for sufficient moisture during key phenological stages, and the temperature threshold for vernalization of winter wheat. SIGNIFICANCEThe integration of optimality concepts for simulating GPP with climate constraints on phenology provides realistic predictions of wheat type and sowing dates. The model thus provides a basis for predicting how crop calendars might change under future climate change. It can also be used to investigate potential changes in management to mitigate the negative impacts of climate change.

POTENCIAL USE OF LOCAL WISEDOM OF SASAK (WARIGE) AS SEASONAL RAINCAST IN LOMBOK

Climate anomaly has pronounced impact on agriculture in Lombok. Occurence of drought results in crop failure. Up to date our knowledge about the cause of climate anomaly has still vey limited. Better understanding on that matter may contribute better water management and cropping strategy. The aim of this study was to examine the skill of seasonal climate forecast based on Sasak local native knowlledge (warige) which was validated with real time planting practised by farmers in different types of climate in Lombok. The output of this study is expected to be used as additional climate information for integrated crop calendar in Lombok. This study applied descriptive methods with survey technique. Data collected included Luni-solar calendar from 1950 to 2012, monthly rainfall data of 1950 to 2012 from four types of climate in Lombok. Five year daily rainfall in selected areas. To obtain moonphase at overhead sun from 1950 to 2012 it was used Kalender Jawa obtained from http:// www.kalendarjawa.com. Ten points of filed crop observation was established in each tipe of climate (C3); D3 and D4. C3 belongs to technical irrigation, D3 belongs to semi-technical irrigation and and D4 belongs to rainfed adn embung irrigation. All farmers involved in this study were also questioned whether they understand and apply warige in their seasonal crop planning. Results of this studyshowed that SCF based on Sasak local native knowledge had good skill in forecasting wetness and dryness of dry season. The different in amount of rainfall in dry and wet years is fairly significant, but usually wet year rainfall did not differ to moderate years’, neither moderate years’ to dry years’. Nevertheless warige was able to predict distribution of rainfall within month during the dry season. Thus we can predict wether it will enough rainfall for several in coming months

Factors affecting adoption intensity of climate change adaptation practices: A case of smallholder rice producers in Chitwan, Nepal

This study examines how smallholder rice producers' adoption intensity for climate change adaptation practices (i.e., improved varieties, irrigation practices, direct seeded rice, integrated pest management, and adjustment in crop calendar) is influenced by access to Extension services, training, weather-related information, and membership in farmer groups or cooperatives (referred to as “institutional resources”). We use survey data collected from 359 smallholder rice producers in the Chitwan district of Nepal in 2019. The results indicate that: (1) access to institutional resources significantly enhance the likelihood of adoption of more climate change adaptation practices; (2) high intensity climate change adaptation practice measured by the adoption of three, four, and five practices significantly increases with access to institutional resources; (3) intensity of adoption of climate change adaptation practices is reduced with greater adaptation alternatives available to rice producers; and (4) lack of information and technical knowledge are the most important reasons for non-adoption of climate change adaptation practices by smallholder rice producers. The results are valuable for policy makers and planners to prioritize training opportunities and allocate scarce resources to enhance climate change adaptation and improve sustainability of rice production practices.

Potential impact of climate change on shifting the calendar of rainfed crops produced in Jordan

Climate change alters the usual pattern of rainfall distribution and temperature in the East Mediterranean region. We investigated the impact of climate change on true season onset, and crop calendar for field crops grown in areas with Semi-Arid-Sub-Humid Mediterranean climate. Analysis showed a clear reduction of annual rainfall for the zone with rainfall more than 200 mm, and above 1°C rise in air temperature for all zones. Generally, the rainfall reduction occurred, in descending order, along the South-North direction. A new approach is proposed to identify the true season onset and the end of the effective rainy season. Shifting the true season onset consequently reduces the crop-growing season and hence the yield. Reducing annual rainfall and the length of effective rainfall season, rising air temperature, and the sharp increase in the number of heat waves will have profound negative impacts on both plant growth and yield.

Changes in Rainfall Pattern in Bengawan Solo Sub-Watershed

<p>Rainfed farming is vulnerable to climate variability, which changes rainfall patterns. Rainfall variability disrupts rainfed rice cultivation because a change in rainfall will affect the rice crop calendar. An analysis of long-term trends over a specific area is required to understand rainfall variability. The aim of this study was to assess climate variability in terms of rainfall magnitude and frequency by analyzing spatial and temporal rainfall trends in Bengawan Solo Sub-Watershed as well as the rainfed rice production. Daily rainfall data from 10 rain gauge stations over the sub-watershed area from the years 1975 to 2020 were used. The data was managed and collected by the Bengawan Solo Watershed authority. Pearson, Mann-Kendall, and Sen’s Slope tests were applied to assess the recorded data correlation, rainfall trends, and magnitude of trends into annual, monthly, and 10-day. The findings of the study indicated the spatial and temporal inhomogeneous rainfall pattern for all locations for 10-day, monthly and annual patterns. The mountainous regions at Tawang Mangu and Ngrambe stations tend to experience an upward trend (positive magnitude), while the coastal regions at Nglirip and Bojonegoro stations have a downward trend(negative magnitude). Those trends also confirmed that coastal regions would be drier than mountainous regions in the future. Understanding this rainfall trend can assist with rainfed farming strategic planning.</p>

Assessing the sensitive spectral bands for soybean water status monitoring and soil moisture prediction using leaf-based hyperspectral reflectance

The stability of soybean yields in Brazil is regularly affected by drought periods, and soil management practices are crucial to expanding the water holding capacity of the soil and providing higher levels of moisture during critical periods, which contribute to increasing crop yields, relieving the need for non-agricultural areas to be converted into croplands. The research reported herein aimed to quantitatively monitor the soil moisture of a soybean crop through leaf-based hyperspectral reflectance and suggest a remote sensing-based approach that might assist in identifying soil management zones. A field experiment at the Brazilian Agricultural Research Corporation during 2016/2017, 2017/2018, and 2018/2019 cropping seasons had ten soybean genotypes subjected to four water conditions: irrigated, non-irrigated, and water deficit induced at the vegetative or reproductive stages. The soil of the experimental site is characterized as Udox Oxisol. Leaf reflectance (400–2500 nm) was collected by the spectroradiometer FieldSpec 3 Jr simultaneously with soil moisture (0–20 and 20–40 cm depths) at eleven dates. Data submitted to Principal Component Analysis (PCA) evaluated the clustering of water conditions and which are the most critical spectral wavelengths to characterize the plant water status. The Partial Least Squares Regression (PLSR) was applied to develop a quantitative spectral model to predict soil moisture. The PCA explained over 93% of the spectral variance within each assessment day, and shortwave infrared wavelengths presented a higher contribution to water status clustering. At the cross-validation step, the PLSR presented R2 up to 0.860 and 0.906 (0–20 and 20–40 cm) underperforming when soil moisture showed no significant differences between water conditions. Using samples from all assessment days, PLSR presented R2 = 0.609 and 0.722 (0–20 and 20–40 cm) at the external validation step (RMSE = 2.7 and 1.9, respectively), with a soil moisture range equal to 16–35% and 20–35% at both depths, remarkably outperforming the traditional univariate spectral models. Our results contribute to soil moisture assessment in extensive soybean areas regardless of the stage of crop development and provide a significant contribution since the Brazilian soybean crop calendar might present differences of over 30 days within the same production region. Due to that, soybean plants are rarely at the same phenological stage on a given date in the season.

In-season and dynamic crop mapping using 3D convolution neural networks and sentinel-2 time series

An accurate, frequently updated, automatic and reproducible mapping procedure to identify seasonal cultivated crops is a prerequisite for many crop monitoring activities. Deep learning was demonstrated to be an effective mapping approach already successfully applied to decametric resolution satellite images (like Sentinel-2 data) to produce yearly crop maps. In this framework, algorithm training is performed with ground truth typically consisting of spatially explicit information available after the end of the season (e.g. yearly crop maps and/or farmer declaration for subsidies at parcel level); however, such data (i) does not allow performing in-season prediction, and (ii) does not provide temporal details fundamental to describe a dynamic crop succession and/or to understand crop management (i.e. planting and harvesting). In this paper we present a Deep Neural Network-based approach capable of generating (i) a crop map of the current season at a specific point in time (“In season mapping” conventionally at the end of the current year), along with (ii) all intermediate maps during the season able to describe in near real-time the evolution of crop presence (“Dynamic-mapping” at the temporal granularity of satellite imagery revisiting, e.g., 5 days for Sentinel-2 data). This approach adopts a smart training procedure of a Deep Neural model by exploiting historical satellite data and ground truth. We introduce a method to automatically generate “short-term” ground truth maps (i.e. 5 days reference) starting from the “long-term” ones (i.e. available yearly static reference) and characterizing temporally the different crop presence by performing a phenological analysis of historical time series. The model was trained and validated in Lombardy (North of Italy) exploiting multi-annual authoritative crop maps from 2016 to 2019. Validation was performed both in time (same areas used for training in a different year) and space (different location) for the year 2019. The quantitative error metrics calculation and Spatio-temporal analysis clearly demonstrate that the model can predict in-season crop presence with a generalization capacity over the long-term (yearly maps: OA > 70% and Kappa > 0.64%) and that the short-term predictions (5 days maps) are coherent with the reference information from expert knowledge (local crop calendars). The model can produce dynamically along the season short-term maps with a medium-high crop-specific User Accuracy at the maximum green-up phase (UA > 53% up to 95%). These products are of extreme interest for final users providing information at the peak of plant development that dynamically changes according to the considered crop, the specific location and the investigated season. These results demonstrate that it is possible to produce a crop map early in the season and extract useful additional information such as crop intensity (e.g. double crops presence) and crop dynamics related to different sowing dates.