From Salinity to Decision-Making: Data Every 15 Minutes Protects Agriculture in the Neretva Valley

A Problem That Begins Far from the Screen

In the Neretva Valley, the decision to irrigate is no longer simply a question of water availability. Water may be available, but its quality at a given moment may pose a risk to crops.

During dry periods, freshwater inflow decreases, while the saltwater wedge penetrates farther upstream through the Neretva riverbed. The effects are felt in both surface and groundwater used for agriculture. Salinity increases, plants are exposed to additional stress, and damage often becomes visible only after it has already occurred.

Farmers reported yellowing leaves after irrigation to the Ministry of Agriculture, Forestry and Water Management of the Herzegovina-Neretva Canton (HNC). This situation showed that the problem was not a lack of data itself, but a lack of timely and understandable information.

For a farmer in the field, a series of technical values is not enough. What is needed is a clear answer to a practical question:

Is the water suitable for irrigation at this moment?

Institutional and Environmental Context

MoWaClim is a cross-border project focused on monitoring the interaction between groundwater and surface water and increasing the resilience of agriculture to the impacts of climate change in coastal systems.

The project is being implemented in Croatia and Bosnia and Herzegovina and covers the Neretva Valley and the Hutovo Blato area. The lead partner is the Dubrovnik-Neretva Canton, while implementation involves the University of Split, Faculty of Civil Engineering, Architecture and Geodesy, the Croatian Agency for Agriculture and Food, the Ministry of Agriculture, Forestry and Water Management of HNC, and the Hutovo Blato Nature Park public institution.

This is an area where water management, agricultural production, and biodiversity conservation are directly interconnected. A change in one parameter can affect multiple sectors, institutions, and user groups.

What Did the Institutions Actually Need?

Surface and groundwater monitoring stations were installed in the field to continuously monitor water conditions. The Ministry of Agriculture, Forestry and Water Management of HNC installed two groundwater and one surface-water monitoring station on the Aluvij Čapljina and Aluvij Hutovo Blato water bodies. The Hutovo Blato Nature Park public institution installed additional surface probes on the Krupa River and at the Ustava.

However, the mere presence of sensors does not solve the problem.

Each station may use different equipment, its own communication protocol, data structure, transmission frequency, and method of parameter identification. Without a central platform, data remain technically fragmented and difficult to apply in everyday operations.

Institutions needed a system that could:

• Automatically retrieve data from different monitoring stations
• Consolidate data regardless of equipment manufacturer
• Verify the validity and completeness of every measurement
• Monitor the availability of stations and sensors
• Identify outdated, duplicated, or suspicious values
• Link every measurement to its location and source
• Enable comparison of surface and groundwater
• Display historical trends
• Alert users when permitted thresholds are exceeded
• Convert technical values into understandable indicators
• Preserve data over the long term for expert analysis and planning.

This was the task that led to the development of the AURA platform.

AURA: From Distributed Devices to a Unified Operational System

AURA IoT is an ITO platform for connecting IoT devices, monitoring stations, environmental data, and expert processes. Its specialized implementation for the MoWaClim project was developed as Hydro Agro Cloud. More information about the platform's capabilities, architecture, and application areas is available on the official website aura-iot.app.

The platform was not designed as yet another isolated dashboard. Its purpose is to establish a complete data flow, from the moment a measurement is generated in the field to the information on which a user can base a decision.

The process consists of several connected steps:

1. The monitoring station records sensor values.
2. Data are automatically delivered to the central platform.
3. AURA identifies the station, sensor, parameter, and measurement time.
4. The system normalizes the data into a common structure.
5. Validation rules check the range, completeness, and logical correctness of the values.
6. The measurement is stored together with source and quality information.
7. Business rules convert values into indicators and alerts.
8. The user is presented with the current status, historical trend, and, where applicable, an operational recommendation.

In this way, AURA connects physical infrastructure, data, expert rules, and the people responsible for resource management.

Data Every 15 Minutes

One of the key requirements of the project was to enable continuous and timely monitoring of water conditions.

Monitoring stations record and transmit data every 15 minutes on parameters such as:

• Salinity
• Electrical conductivity
• Water temperature
• pH
• Dissolved oxygen
• Oxygen saturation
• Nitrates
• Other water quality indicators.

The measurement frequency is important because water conditions can change throughout the day depending on flow, temperature, rainfall, drought conditions, and seawater intrusion.

Occasional laboratory measurement can confirm that a problem existed at a particular moment. Continuous digital monitoring makes it possible to track when the problem began, how long it lasted, how it developed, and when conditions became acceptable again.

From a Technical Value to an Irrigation Recommendation

One of the system's most important functions is converting complex measurements into simple and understandable information.

Electrical conductivity and salinity can be presented as precise numerical values for expert users. A farmer, however, needs an immediate interpretation of those values.

AURA therefore enables the calculation of water suitability indicators for irrigation. Depending on the measured values and defined expert thresholds, the condition can be displayed through clear categories:

• Safe for irrigation
• Caution required
• Irrigation is not recommended.

Each indicator also includes the time of the latest measurement. If a station has not provided a new reading for a certain period, the system clearly marks this so that the user does not make a decision based on outdated information.

This is an important distinction between a system that merely displays data and a platform that supports decision-making.

Data Quality Monitoring

An operational decision can only be reliable if the data on which it is based are reliable.

For this reason, AURA checks every received measurement. The system applies rules that can identify:

• A value outside the permitted range
• An illogical timestamp
• A duplicate measurement
• Missing data
• An unexpected communication interruption
• A value that significantly deviates from previous measurements
• Data that are technically valid but require expert review.

Suspicious data do not have to be deleted automatically. They are marked with an appropriate quality status, allowing expert users to analyze them later.

Alongside the measurements themselves, the system records the data source, the time of receipt, the communication protocol, and other available metadata. This ensures traceability from the displayed indicator back to the original sensor record.

Central Monitoring of Stations and Sensors

A monitoring system must enable the monitoring of its own infrastructure, not just the environment.

Through the AURA platform, administrators and operators can monitor:

• Whether a station is active
• The time of the last received data
• The expected measurement frequency
• The number of received records
• The status of individual sensors
• Availability history
• Active alerts
• Failed integration messages
• Suspicious or invalid values.

If a station stops transmitting data, the system can automatically detect the interruption and alert the responsible person. This allows a technical problem to be identified before it causes a larger gap in the data time series.

Connecting Automated and Expert Data

Some project data are collected automatically through sensors, but a complete analysis of environmental and agricultural conditions also requires data collected by experts in the field and in laboratories.

AURA therefore supports structured manual entry of:

• Measurements using mobile probes
• Mandarin phenological stages
• Laboratory fruit analyses
• Biological monitoring
• Expert observations and supporting data.

Manual entries are not kept in separate documents or spreadsheets. Data are linked to location, period, user, and other measurements within the system.

The platform supports field validation, saving incomplete entries, subsequent completion, expert review, approval, comments, and a complete history of changes.

This brings automated sensor measurements and the institutions' expert knowledge together on a shared data foundation.

Interactive Map and Historical Trends

Location is essential for understanding environmental data.

AURA displays monitoring stations on an interactive map of the pilot area. Users can select a location and review:

• The current station status
• Latest values
• The time of the latest measurement
• Active alerts
• Available parameters
• Trends over a selected period.

Expert users can compare multiple stations and parameters. This makes it possible to analyze the relationship between groundwater and surface water, monitor the spread of salinity, and compare changes with temperature, rainfall, and other environmental indicators.

Long-term data collection provides a foundation for scientific analysis, agricultural production planning, and more responsible management of water resources.

Architecture Independent of a Single Vendor

One of the fundamental architectural principles of the AURA platform is independence from equipment manufacturers.

The platform supports multiple methods of data ingestion, including HTTP, MQTT, and file transfer via FTP or SFTP protocols. Different input formats are converted into a common data model.

The configuration of stations, sensors, measurement units, validation ranges, and alert thresholds is stored within the platform itself. A new station can be configured through the administration interface without developing a separate application for each type of equipment.

This approach enables institutions to:

• Use existing monitoring equipment
• Combine devices from different manufacturers
• Gradually expand the sensor network
• Reduce dependence on a single supplier
• Preserve previous investments in infrastructure.

Technical Foundation of the System

AURA was developed as a modular and scalable platform designed to work with large volumes of timestamped measurements.

The technical foundation includes:

• .NET 8 backend and REST API
• React user and administration interface
• PostgreSQL database
• TimescaleDB for time-series data
• MQTT communication with IoT devices
• HTTP and file-based integration services
• Docker containerization
• Automated backups
• Centralized logging
• Technical monitoring and metrics.

Data are organized by institutions, monitoring stations, sensors, parameters, and locations. Multiple organizations can use the same platform with logical separation of users, configurations, and data.

The architecture allows the system to be further expanded with new stations, sensor types, partners, and application areas.

Security and Access Management

The project involves multiple institutions and different categories of users. Therefore, access management is built into the platform's core architecture.

User permissions are determined by organization, role, and assigned privileges. The system distinguishes between platform administrators, institutional administrators, expert users, operators, and public users.

All important administrative and data-related activities are recorded through audit logs. It is possible to determine who made a change, when the change occurred, and which data it concerned.

The public part of the system publishes only selected indicators and aggregated data. Administrative functions, expert data, and technical configurations are available exclusively to authorized users.

Results Achieved

By implementing the AURA platform within the MoWaClim project, a shared digital foundation was established for continuous monitoring of environmental and agricultural parameters.

Key results include:

• Centralized collection of data from surface and groundwater monitoring stations
• Availability of new measurements at 15-minute intervals
• Integration of different devices and data formats
• Automatic measurement quality checks
• Monitoring of station and sensor status
• Interactive geographic overview of monitoring locations
• Display of current values and historical trends
• Integration of automated, field, and laboratory data
• Timely information for farmers about water conditions
• A long-term data foundation for institutions managing water resources and agriculture
• An architecture that can be extended to new devices, locations, and projects.

The most important result is not simply a greater amount of available data. The result is a shorter time between a change in the field, its detection, and the user's response.

Why Does This Matter?

Seawater intrusion cannot be stopped by a software platform. It is possible, however, to detect a change early enough, understand its intensity, and help the user avoid a decision that could further endanger the crop.

This is precisely the role of the AURA platform.

For a farmer, it means having the information needed before turning on the irrigation system.

For an agronomist, it means the ability to compare data across locations and periods.

For a technical operator, it means timely detection of a station that has stopped transmitting data.

For an institution, it means a reliable record on which it can plan climate change adaptation measures.

For the scientific community, it means structured, long-term time-series data.

Technology does not replace expert knowledge in this case. It ensures that quality data reach the responsible person at a moment when they can still influence a decision.

Beyond a Single Implementation

Hydro Agro Cloud represents a specialized implementation of the AURA platform for the needs of the MoWaClim project. AURA itself is designed for a much broader range of applications.

The same architectural foundation can be used for:

• Early flood warning
• Water level and water quality monitoring
• Air quality monitoring
• Monitoring fire and environmental risks
• Management of agricultural sensors
• Monitoring of utility infrastructure
• Smart cities
• Industrial monitoring
• Energy
• Critical infrastructure management.

Parameters, devices, and expert rules vary depending on the application area, but the core process remains the same: connect sources, verify data, identify risk, and enable timely decision-making.

Engineering Takeaway

This project confirms one of the fundamental principles of the ITO approach to digital transformation: value does not come simply from installing sensors or creating yet another user interface.

Value is created when physical equipment, communication protocols, data, expert rules, and institutional responsibilities are connected into a unified operational system.

AURA was developed precisely according to this principle. Instead of tying an institution to a single manufacturer or closed technical ecosystem, the platform establishes a neutral integration and data layer. Existing devices can be retained, new ones can be gradually integrated, and the institution retains control over users, rules, and data.

In the MoWaClim project, this approach transforms salinity measurement into information that can protect crops. In other domains, the same principle can mean earlier warning of floods, infrastructure failures, pollution, or other operational risks.

Trust Principles Applied

This project was developed according to the following principles:

✅ Security by Design

✅ Privacy by Design

✅ Interoperability

✅ Auditability

✅ Open Architecture

✅ Vendor Independence

✅ Data Quality by Design

✅ Institution-controlled Governance

From Sensors to Timely Decision-Making

AURA enables institutions to connect different sensors, monitoring stations, and data sources in a unified operational environment. The platform adapts to the specific application area, existing equipment, and institutional rules, from environmental monitoring and agriculture to civil protection, smart cities, and critical infrastructure.

Discover the AURA IoT platform or visit aura.ito.ba for more information about implementation possibilities.