A proposal to train young people to investigate risk in their own neighbourhoods. From schoolyards to headwaters, students would study rainfall with teachers and technicians, test sensors and discover careers in disaster prevention.
The proposed pilot begins in three municipalities facing landslides and flash floods. A national coverage study would guide later expansion to places with few observations.
Cássio’s proposal connects six strands. A school observation could lead to a research question, a prototype or a technical course. Each strand has a practical result to test.
Teachers and students would investigate rainfall, past disasters and local evacuation routes. Science clubs and drills would bring families into the conversation.
Students would compare instruments safely at school. Technical teams would install and maintain sensors in the headwaters, on slopes and in watercourses.
Students could help recover historical accounts. Specialists would align event times and locations with rainfall and soil records to study short-duration rainfall and slope stability.
Specialists and civil defence teams would test local thresholds and document hits and false alarms. Official warnings and emergency decisions remain with the competent authorities.
A real failure, such as losing a signal during a storm, could become a challenge for students and technical institutes. A validated prototype could enter incubation or intellectual property review.
Technical schools and federal institutes could teach instrumentation, hydrology and geotechnics. Supervised placements would connect the courses to civil defence and local maintenance.
Choose a strand to see what it would deliver.
Read the full project sheetChoose a point on the diagram. The question and the location of the process determine which instrument belongs there.
A recording rain gauge tracks when rain starts and how its intensity changes. A storm near the headwaters can be hard to detect from a gauge in the valley.
Soil moisture sensors and piezometers help track infiltration and changes in water pressure. Geotechnical specialists interpret these readings alongside site conditions.
A water-level sensor records the passage of a flood. Its timing can be compared with upstream rainfall and with documented local events.
Automatic and manual rain gauges become part of the same lesson. Students investigate differences at a safe site and share the findings with their families.
Illustrative animation with schematic curves, not measured data. It shows why instrument placement matters.
Cássio proposes Nova Friburgo, Petrópolis and São Sebastião as candidates. Brazil–Japan cooperation through the SABO project offers a technical foundation in all three areas. Participation by education authorities and municipalities still requires agreement.
Relate headwater rainfall to landslide scars from 2011 and use the instrument readings in a local technical school.
Compare hillslope rain with water levels in urban rivers and practise communicating with downstream neighbourhoods alongside civil defence.
Investigate how debris flows from the Serra do Mar reach valley outlets and how families identify safe evacuation routes.
Education departments would match enrolment information to high- and very-high-risk areas internally. The project would receive counts by school and risk area, protecting student names and addresses. Priority would follow the students who live in those areas, even when they attend school elsewhere.
Federal monitoring was present in all three candidates in the August study. The pilot asks where measurements are needed within a catchment. For later expansion, the national study of 5,570 municipalities identifies another need: places outside the networks.
municipalities had no rain station identified in ANA, Cemaden or INMET records.
8.42 million residents in these municipalities · 2025 estimate
municipalities lacked operational monitoring under the study criteria. This includes the 1,042 with no station.
18.93 million residents in these municipalities · 2025 estimate
The snapshot is dated 25 August 2026. Station presence includes ANA’s rainfall inventory, including historical stations, Cemaden’s registered automatic rain gauges and INMET’s automatic weather stations.
Operational coverage is approximated by an ANA observation since 2021, a rain gauge in Cemaden’s register, or a station marked operational by INMET.
The study uses 2022 municipal boundaries. A nearby gauge across the boundary counts for the neighbouring municipality. The analysis excludes state, municipal and private networks, radar and satellites. Each dot represents a municipality. The table includes islands outside the map frame. Assessing risk or choosing a school requires local coverage, disaster records and exposure information.
| State | Assessed | No station | Operational gap |
|---|---|---|---|
| AC | 22 | 1 | 1 |
| AL | 102 | 21 | 41 |
| AM | 62 | 0 | 0 |
| AP | 16 | 0 | 0 |
| BA | 417 | 50 | 166 |
| CE | 184 | 0 | 0 |
| DF | 1 | 0 | 0 |
| ES | 78 | 0 | 0 |
| GO | 246 | 102 | 133 |
| MA | 217 | 50 | 71 |
| MG | 853 | 222 | 325 |
| MS | 79 | 6 | 7 |
| MT | 141 | 25 | 34 |
| PA | 144 | 25 | 36 |
| PB | 223 | 1 | 2 |
| PE | 185 | 0 | 2 |
| PI | 224 | 55 | 139 |
| PR | 399 | 44 | 108 |
| RJ | 92 | 3 | 5 |
| RN | 167 | 30 | 60 |
| RO | 52 | 19 | 22 |
| RR | 15 | 1 | 1 |
| RS | 497 | 172 | 214 |
| SC | 295 | 74 | 85 |
| SE | 75 | 22 | 53 |
| SP | 645 | 55 | 212 |
| TO | 139 | 64 | 72 |
The proposed station would keep measuring overnight and during school holidays. During term, teachers and students could investigate the rainfall record from their own school.
An automatic rain gauge would send readings to the proposed platform. With a teacher’s guidance, students could check the total against a manual gauge and investigate any differences.
A science club could build a second instrument, the Shadow Station. Comparing the two instruments gives students a practical reason to learn about calibration, charts and measurement uncertainty.
A school bulletin could connect rainfall records with residents’ flood memories. Under a proposed observation protocol, adults would agree how to share information with local civil defence. Official warnings remain the responsibility of the competent authorities.
If all the water falling on one flat square metre were collected, each millimetre would yield one litre. Move the slider to compare rainfall depth and volume per unit area.
The draft proposes 56 hours of training for at least two teachers per school. Education authorities would need to give staff time to plan lessons, guide the science club and prepare colleagues to take over the following year.
Adults would check that the gauge is level, the power supply works and transmission gaps are recorded. Before installation, the school and technical team would assign responsibility for faults, spare parts and holiday coverage.
The proposed platform would organise each school’s time series, flag missing or suspect readings and export data for lessons. Each observation would need a unit, time, interval, station identifier and quality flag.
The project sheet identifies the Hunan seminar of 2 November 2026 as the starting point for this discussion. Technical schools and researchers in both countries could investigate the same measurement problem.
Cemaden Education’s Pluvipet + App activities guide schools and residents in measuring rain with low-cost instruments and discussing what the data reveal about their surroundings.
Cemaden Education ↗Shanghai No. 4 Middle School has a rain gauge at its weather station, which connects to the municipal meteorological data platform. Students’ records are used in geography and physics classes.
Xuhui District Education Bureau ↗One technical school in each country could compare a manual reading with an automatic one, simulate a transmission failure and document data recovery. A calibration record, a sample time series and a bilingual lesson would give both teams evidence for deciding whether to extend the exchange.
Scholarships, equipment supply and agreements would need negotiation through official channels, including Brazil’s foreign ministry and cooperation agency.
Choose a milestone to explore the targets proposed in the project sheet. The initial phase would establish the baseline, costs and delivery capacity of schools and municipalities.
| Proposed indicator | 2028 | 2030 | 2034 |
|---|---|---|---|
| Participating schools | 100 | 180 | 300 |
| Participating young people | 5,000 | 12,000 | 30,000 |
| Teachers trained | 200 | 500 | 1,000 |
| Technicians trained | 150 | 600 | 1,500 |
| Specialists trained | 20 | 60 | 150 |
| Scholarship holders in China (courses, degrees, postgraduate) | 1–2 | 3–5 | 5–10 |
| Active monitoring stations | 100 | 250 | 500 |
| Catchments with their own warning trigger | 3 | 8 | 20 |
| Patent applications and software registrations | 1 | 4 | 10 |
| Incubated startups | 2 | 6 | 15 |
| Reduction in deaths (vs. 2011–2026 average) | — | −25% | −50% |
| Reduction in damage in events with warning | — | −15% | −30% |
For landslide and flash-flood deaths and damage in events with advance warning, the proposed baseline is the 2011–2026 average in participating municipalities. Rainfall severity varies between years and must be considered when estimating the project's effect.
The project sheet asks the department to consider the concept and decide whether to begin formal discussions. The dates describe a proposed sequence. Scope, funding, responsibilities and agreements still need to be settled.
DPM and CGNAT would prepare a project charter, consult science and education ministries and municipalities, identify schools from aggregate data and cost maintenance and training.
After participation is agreed, start clubs, install and calibrate instruments, and test a first lesson. Check thresholds against recorded events and refine drills.
Compare transmission during storms, classroom use, warning performance and technical training outcomes. The partners could then judge whether the approach can transfer to other catchments.
Cássio’s trilingual project sheet dated 2 November 2026 sets out the strands, pilot and targets. DPM’s study of 25 August 2026 uses ANA, Cemaden and INMET records, IBGE’s 2022 boundaries and 2025 population estimates from IPS Brasil 2026. The Cunha Porã example draws on Brazil’s Digital Disaster Atlas. Municipality names in the download were checked against IBGE in November 2026.