Luciana Alvim Santos Romani
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Luciana Alvim Santos Romani has a total of 77 co-authors. Here are the top co-authors by number of shared publications:
- Silvia Maria Fonseca Silveira Massruhá
Embrapa, W3 Norte Avenue-Asa Norte, Brasilia 70770-901, DF, Brazil
10 shared publications - Édson Luis Bolfe
Brazilian Agricultural Research Corporation, Embrapa Digital Agriculture, Campinas 13083-886, São Paulo, Brazil
5 shared publications - Jurandir Zullo Junior
Universidade Estadual de Campinas
4 shared publications - Ivan Bergier
Embrapa Digital Agriculture, Brazilian Agricultural Research Corporation, Campinas 13083-886, SP, Brazil
3 shared publications - Jayme Garcia Arnal Barbedo
Embrapa Digital Agriculture, Embrapa, 209 André Tosello Avenue, Campinas 13083-886, SP, Brazil
3 shared publications - João Paulo da Silva
Embrapa
3 shared publications
Latest Publications
- Open Access
- 19 November 2026
- Published by MDPI AG in Sustainability
- Vol. 18 (22) , 11791
Hydroclimatic Variability and Irrigation Planning Using CropWaterBalance: A Case Study in Brazil
Effective irrigation management is essential for adapting agricultural production to climate change. This study investigates hydroclimatic variability across five Agrotechnological Districts (DATs) in Brazil and estimates irrigation water requirements under current and projected climatic conditions. Meteorological data from CPTEC/INPE (2000–2025) were analyzed using the Mann–Kendall and Pettitt tests to identify monotonic trends and abrupt change points, respectively. Nonstationarity was assessed using Generalized Additive Models for Location, Scale, and Shape (GAMLSS), while irrigation requirements were simulated using the CropWaterBalance R package. The Mann–Kendall test detected a significant increase in maximum air temperature in Alto Alegre and a significant decrease in precipitation in Caconde, with no significant monotonic trends in the remaining climatic series. Additionally, GAMLSS selected nonstationary models for minimum air temperature in all five DATs and maximum air temperature in two, suggesting temporal changes in climatic distributions not necessarily reflected in significant monotonic trends. The two crop–location combinations with the highest simulated irrigation water requirements, sugarcane in Alto Alegre and coffee in Caconde, were selected for future climate simulations using bias-corrected CMIP6 projections from the CLIMBra dataset under the SSP5-8.5 scenario. For 2030, simulated irrigation water requirements increased by 32.6% for sugarcane and 8.2% for coffee relative to 2025 under the selected climate projection. Overall, the findings reveal spatial differences in hydroclimatic variability and their potential implications for irrigation water requirements, providing information that may support future irrigation planning and climate adaptation assessments.
- 19 November 2026
- Published by MDPI AG in Sustainability
- Vol. 18 (22) , 11791
Open AccessHydroclimatic Variability and Irrigation Planning Using CropWaterBalance: A Case Study in Brazil
Effective irrigation management is essential for adapting agricultural production to climate change. This study investigates hydroclimatic variability across five Agrotechnological Districts (DATs) in Brazil and estimates irrigation water requirements under current and projected climatic conditions. Meteorological data from CPTEC/INPE (2000–2025) were analyzed using the Mann–Kendall and Pettitt tests to identify monotonic trends and abrupt change points, respectively. Nonstationarity was assessed using Generalized Additive Models for Location, Scale, and Shape (GAMLSS), while irrigation requirements were simulated using the CropWaterBalance R package. The Mann–Kendall test detected a significant increase in maximum air temperature in Alto Alegre and a significant decrease in precipitation in Caconde, with no significant monotonic trends in the remaining climatic series. Additionally, GAMLSS selected nonstationary models for minimum air temperature in all five DATs and maximum air temperature in two, suggesting temporal changes in climatic distributions not necessarily reflected in significant monotonic trends. The two crop–location combinations with the highest simulated irrigation water requirements, sugarcane in Alto Alegre and coffee in Caconde, were selected for future climate simulations using bias-corrected CMIP6 projections from the CLIMBra dataset under the SSP5-8.5 scenario. For 2030, simulated irrigation water requirements increased by 32.6% for sugarcane and 8.2% for coffee relative to 2025 under the selected climate projection. Overall, the findings reveal spatial differences in hydroclimatic variability and their potential implications for irrigation water requirements, providing information that may support future irrigation planning and climate adaptation assessments.
- Open Access
- 15 September 2026
- Published by Elsevier in SSRN Electronic Journal
Beyond Connectivity: The Practice-Capability-Knowledge Nexus in Digital Agriculture in the Brazilian Caatinga
The digitalisation of agriculture is frequently presented as a means of increasing productivity and including family farmers. Its incorporation into rural territories, however, depends on material, cultural, institutional, and epistemic conditions that extend beyond infrastructure. This study analyses how Technical Assistance and Rural Extension (ATER) professionals interpret these conditions in Boa Vista do Tupim, in the Brazilian Caatinga. A focus group was examined through qualitative content analysis at the manifest and latent levels, together with analysis of verbal sequences. The results indicate constraints related to certification, production planning, logistics, uneven connectivity, and the management of collective resources. The interactional sequences show how explanations were qualified or expanded during discussions of informal commercial agreements, uses of connectivity, and collective financing instruments. Social Practice Theory, the Capability Approach, and the Ecology of Knowledges are integrated through the Practice-Capability-Knowledge Nexus, whose components appear with unequal intensity across the three empirical dimensions. The study indicates that digital technologies need to be examined in relation to everyday practices, the conditions under which resources are converted into opportunities for action, and the forms of knowledge recognised in the territory, rather than as resources whose provision alone ensures incorporation.
- 15 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessBeyond Connectivity: The Practice-Capability-Knowledge Nexus in Digital Agriculture in the Brazilian Caatinga
The digitalisation of agriculture is frequently presented as a means of increasing productivity and including family farmers. Its incorporation into rural territories, however, depends on material, cultural, institutional, and epistemic conditions that extend beyond infrastructure. This study analyses how Technical Assistance and Rural Extension (ATER) professionals interpret these conditions in Boa Vista do Tupim, in the Brazilian Caatinga. A focus group was examined through qualitative content analysis at the manifest and latent levels, together with analysis of verbal sequences. The results indicate constraints related to certification, production planning, logistics, uneven connectivity, and the management of collective resources. The interactional sequences show how explanations were qualified or expanded during discussions of informal commercial agreements, uses of connectivity, and collective financing instruments. Social Practice Theory, the Capability Approach, and the Ecology of Knowledges are integrated through the Practice-Capability-Knowledge Nexus, whose components appear with unequal intensity across the three empirical dimensions. The study indicates that digital technologies need to be examined in relation to everyday practices, the conditions under which resources are converted into opportunities for action, and the forms of knowledge recognised in the territory, rather than as resources whose provision alone ensures incorporation.
- 2 September 2026
- Published by Springer Nature
- p. 293-305
Science Center for Development in Digital Agriculture
- 2 September 2026
- Published by Springer Nature
- p. 293-305
Science Center for Development in Digital Agriculture
- Open Access
- 18 August 2026
- Published by Elsevier in Remote Sensing Applications: Society and Environment
- Vol. 43
Seasonal behavior and yield penalty classification in pasture areas based on SAR polarimetry and machine learning techniques
- 18 August 2026
- Published by Elsevier in Remote Sensing Applications: Society and Environment
- Vol. 43
Open AccessSeasonal behavior and yield penalty classification in pasture areas based on SAR polarimetry and machine learning techniques
- Open Access
- 1 July 2026
- Published by Elsevier in Smart Agricultural Technology
- Vol. 14
Diagnostic framework for assessing Smart Agricultural Technology adoption
- 1 July 2026
- Published by Elsevier in Smart Agricultural Technology
- Vol. 14
Open AccessDiagnostic framework for assessing Smart Agricultural Technology adoption
- Open Access
- 1 May 2026
- Published by Wiley in Agronomy Journal
- Vol. 118 (3)
Thermal sensitivity and yield of Arabica coffee under climatic variability
Coffee yield is influenced by climatic factors, with temperature and water availability being key determinants. This study aimed to evaluate the effects of climate variability on the yield of Arabica coffee ( Coffea arabica L.) cultivars across phenological phases and to validate an agrometeorological model at local and regional scales using yield–climate time‐series data from the Volcanic Region of Poços de Caldas. The Arabica coffee cultivars Mundo Novo 376/4, Catuaí IAC 144, and Bourbon Vermelho were evaluated across four phenological phases. Pearson's correlation coefficients were applied for each phenological phase and production year between yield and climatic variables, along with the Mann–Kendall test to detect monotonic trends in climatic data. The agrometeorological model was evaluated at different analytical scales and under distinct cultivars and yield scenarios, using historical data from 2011 to 2024. The grain‐filling phase was the most sensitive to increasing temperature, while flowering showed a rising warming trend, increasing its vulnerability. Cultivars differed in their thermal and water tolerance, with Bourbon Vermelho showing high vegetative resilience but marked reproductive susceptibility. The model performed well at the regional scale ( R 2 = 0.93; RMSE = 453 kg ha −1 ), particularly within the Volcanic Region, but exhibited limited accuracy at the plot scale due to local variability. Estimated yield losses were mainly associated with water deficit, followed by frost events, while thermal penalties were minimal. These results highlight the need to improve climate‐resilience strategies and mitigate the impacts of climate change on coffee cultivation.
- 1 May 2026
- Published by Wiley in Agronomy Journal
- Vol. 118 (3)
Open AccessThermal sensitivity and yield of Arabica coffee under climatic variability
Coffee yield is influenced by climatic factors, with temperature and water availability being key determinants. This study aimed to evaluate the effects of climate variability on the yield of Arabica coffee ( Coffea arabica L.) cultivars across phenological phases and to validate an agrometeorological model at local and regional scales using yield–climate time‐series data from the Volcanic Region of Poços de Caldas. The Arabica coffee cultivars Mundo Novo 376/4, Catuaí IAC 144, and Bourbon Vermelho were evaluated across four phenological phases. Pearson's correlation coefficients were applied for each phenological phase and production year between yield and climatic variables, along with the Mann–Kendall test to detect monotonic trends in climatic data. The agrometeorological model was evaluated at different analytical scales and under distinct cultivars and yield scenarios, using historical data from 2011 to 2024. The grain‐filling phase was the most sensitive to increasing temperature, while flowering showed a rising warming trend, increasing its vulnerability. Cultivars differed in their thermal and water tolerance, with Bourbon Vermelho showing high vegetative resilience but marked reproductive susceptibility. The model performed well at the regional scale ( R 2 = 0.93; RMSE = 453 kg ha −1 ), particularly within the Volcanic Region, but exhibited limited accuracy at the plot scale due to local variability. Estimated yield losses were mainly associated with water deficit, followed by frost events, while thermal penalties were minimal. These results highlight the need to improve climate‐resilience strategies and mitigate the impacts of climate change on coffee cultivation.
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23, 2505805
Elevation-Based Clustering and Spatiotemporal Analysis of Coffee Agro-Environmental Dynamics
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23, 2505805
Open AccessElevation-Based Clustering and Spatiotemporal Analysis of Coffee Agro-Environmental Dynamics
- Open Access
Hail Netting in Apple Orchards: Current Knowledge, Research Gaps, and Perspectives for Digital Agriculture
Hailstorms are a major climatic threat to apple production, causing substantial economic losses in orchards worldwide. Anti-hail nets have been increasingly adopted to mitigate this risk, but the scientific literature on their effectiveness and future applications remains scattered, especially considering advances in digital agriculture. This study synthesizes current knowledge on the use of anti-hail nets in apple orchards through a systematic review and explores future perspectives involving digital technologies. A PRISMA-based review was conducted using three databases, revealing information regarding the studied countries, netting colors, and apple varieties, among others. A clear research gap was identified in integrating anti-hail nets with remote sensing and Artificial Intelligence (AI). This paper also analyzes studies from Vacaria, Brazil, a key apple-producing region and part of the Semear Digital project, highlighting local efforts to use hail netting in commercial orchards. Potential applications of AI algorithms and remote sensing are proposed for hail netting assessment, orchard monitoring, and decision-making support. These technologies can improve predictive modeling, quantify areas, and enhance precision management. Findings suggest combining traditional protective methods with technological innovations to strengthen orchard resilience in regions exposed to extreme weather.
Open AccessHail Netting in Apple Orchards: Current Knowledge, Research Gaps, and Perspectives for Digital Agriculture
Hailstorms are a major climatic threat to apple production, causing substantial economic losses in orchards worldwide. Anti-hail nets have been increasingly adopted to mitigate this risk, but the scientific literature on their effectiveness and future applications remains scattered, especially considering advances in digital agriculture. This study synthesizes current knowledge on the use of anti-hail nets in apple orchards through a systematic review and explores future perspectives involving digital technologies. A PRISMA-based review was conducted using three databases, revealing information regarding the studied countries, netting colors, and apple varieties, among others. A clear research gap was identified in integrating anti-hail nets with remote sensing and Artificial Intelligence (AI). This paper also analyzes studies from Vacaria, Brazil, a key apple-producing region and part of the Semear Digital project, highlighting local efforts to use hail netting in commercial orchards. Potential applications of AI algorithms and remote sensing are proposed for hail netting assessment, orchard monitoring, and decision-making support. These technologies can improve predictive modeling, quantify areas, and enhance precision management. Findings suggest combining traditional protective methods with technological innovations to strengthen orchard resilience in regions exposed to extreme weather.
- Open Access
- 12 September 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (18) , 3168
Dense Time Series of Harmonized Landsat Sentinel-2 and Ensemble Machine Learning to Map Coffee Production Stages
Coffee demand continues to rise, while producing countries face increasing challenges and yield losses due to climate change. In response, farmers are adopting agricultural practices capable of boosting productivity. However, these practices increase intercrop variability, making coffee mapping more challenging. In this study, a novel approach is proposed to identify coffee cultivation considering four phenological stages: planting (PL), producing (PR), skeleton pruning (SK), and renovation with stumping (ST). A hierarchical classification framework was designed to isolate coffee pixels and identify their respective stages in one of Brazil’s most important coffee-producing regions. A dense time series of multispectral bands, spectral indices, and texture metrics derived from Harmonized Landsat Sentinel-2 (HLS) imagery, with an average revisit time of ~3 days, was employed. This data was combined with an ensemble learning approach based on decision-tree algorithms, specifically Random Forest (RF) and Extreme Gradient Boosting (XGBoost). The results achieved unprecedented sensitivity and specificity for coffee plantation detection with RF, consistently exceeding 95%. The classification of coffee phenological stages showed balanced accuracies of 77% (ST) and from 93% to 95% for the other classes. These findings are promising and provide a scalable framework to monitor climate-resilient coffee management practices.
- 12 September 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (18) , 3168
Open AccessDense Time Series of Harmonized Landsat Sentinel-2 and Ensemble Machine Learning to Map Coffee Production Stages
Coffee demand continues to rise, while producing countries face increasing challenges and yield losses due to climate change. In response, farmers are adopting agricultural practices capable of boosting productivity. However, these practices increase intercrop variability, making coffee mapping more challenging. In this study, a novel approach is proposed to identify coffee cultivation considering four phenological stages: planting (PL), producing (PR), skeleton pruning (SK), and renovation with stumping (ST). A hierarchical classification framework was designed to isolate coffee pixels and identify their respective stages in one of Brazil’s most important coffee-producing regions. A dense time series of multispectral bands, spectral indices, and texture metrics derived from Harmonized Landsat Sentinel-2 (HLS) imagery, with an average revisit time of ~3 days, was employed. This data was combined with an ensemble learning approach based on decision-tree algorithms, specifically Random Forest (RF) and Extreme Gradient Boosting (XGBoost). The results achieved unprecedented sensitivity and specificity for coffee plantation detection with RF, consistently exceeding 95%. The classification of coffee phenological stages showed balanced accuracies of 77% (ST) and from 93% to 95% for the other classes. These findings are promising and provide a scalable framework to monitor climate-resilient coffee management practices.
- Open Access
- 5 May 2025
- Published by RGSA- Revista de Gestao Social e Ambiental in Revista de Gestão Social e Ambiental
- Vol. 19 (5) , e012197
Characterization and Efficiency of Beef Cattle Production Systems in Mato Grosso do Sul
Objective: to characterize modal beef cattle systems in municipalities in Mato Grosso do Sul (MS), through the use of unstructured data analytics methodologies aimed at diagnostic analysis, in order to identify intensification trends in the systems. Theoretical Framework: Mato Grosso do Sul (MS) is a significant beef cattle producing region in Brazil, ranking fifth in the country with an inventory of approximately 20 million head. The beef production systems in MS, presents a wide variety of production systems, which have been changing over time. Method: used the representative farms methodology to collect data on 13 beef cattle production systems in MS. We then employed unstructured data analytics methodologies to conduct a diagnostic analysis, with the goal of identifying trends towards intensification in these systems. We estimated the efficiency measures of these different production modes and highlighted those with the best performance. Results and Discussion: Corumbá (cow-calf) and Naviraí (rearing and fattening) systems being allocated to separate groups. This separation is justified by the fact that they are the most extensive and most intensive production systems, respectively. The other two groups were cow-calf systems/full cycle and rearing/fattening systems. In addition, we calculated efficiency scores using a production approach. The average efficiency, calculated using data envelopment analysis models, was 54.2%. A fractional regression fit to the efficiency scores and the types of production systems as covariates suggested that there are statistically significant differences between the production systems. Research Implications: the typical farms on a municipal scale in MS reflect the sustainability factors of beef cattle activity in the state. The managerial implications involve guiding the technical and scientific analysis of the processes. Originality/Value: unstructured machine learning methods successfully synthesized production variables, enabling the classification of systems into clusters of homogeneous municipalities based on practiced livestock production systems. The two-stage DEA methodology used in this analysis helped identify significant contextual variables for beef cattle production.
- 5 May 2025
- Published by RGSA- Revista de Gestao Social e Ambiental in Revista de Gestão Social e Ambiental
- Vol. 19 (5) , e012197
Open AccessCharacterization and Efficiency of Beef Cattle Production Systems in Mato Grosso do Sul
Objective: to characterize modal beef cattle systems in municipalities in Mato Grosso do Sul (MS), through the use of unstructured data analytics methodologies aimed at diagnostic analysis, in order to identify intensification trends in the systems. Theoretical Framework: Mato Grosso do Sul (MS) is a significant beef cattle producing region in Brazil, ranking fifth in the country with an inventory of approximately 20 million head. The beef production systems in MS, presents a wide variety of production systems, which have been changing over time. Method: used the representative farms methodology to collect data on 13 beef cattle production systems in MS. We then employed unstructured data analytics methodologies to conduct a diagnostic analysis, with the goal of identifying trends towards intensification in these systems. We estimated the efficiency measures of these different production modes and highlighted those with the best performance. Results and Discussion: Corumbá (cow-calf) and Naviraí (rearing and fattening) systems being allocated to separate groups. This separation is justified by the fact that they are the most extensive and most intensive production systems, respectively. The other two groups were cow-calf systems/full cycle and rearing/fattening systems. In addition, we calculated efficiency scores using a production approach. The average efficiency, calculated using data envelopment analysis models, was 54.2%. A fractional regression fit to the efficiency scores and the types of production systems as covariates suggested that there are statistically significant differences between the production systems. Research Implications: the typical farms on a municipal scale in MS reflect the sustainability factors of beef cattle activity in the state. The managerial implications involve guiding the technical and scientific analysis of the processes. Originality/Value: unstructured machine learning methods successfully synthesized production variables, enabling the classification of systems into clusters of homogeneous municipalities based on practiced livestock production systems. The two-stage DEA methodology used in this analysis helped identify significant contextual variables for beef cattle production.