Skip to content

NASA Space Apps Challenge · Costa Rica

From orbitto the furrow.

Mission Valdocco brings a farm station's readings, NASA Earth observation and weather forecasts into one view a farmer can read.

8
NASA products to explore
4
sensor groups on the station
7
days of forecast maps
1
simple view in two languages

The problem

The information exists. It is scattered.

Farmers decide while the weather changes, but the useful data lives in separate sensors, maps, forecasts and scientific datasets. They use different units, cover different areas and update at different times. A satellite image or a lone number is hard to interpret.

Our goal is to bring that context together in one place and present it clearly. It supports observation and planning; it does not replace an agronomist, a field inspection or the farmer's judgment.

How it works

Five steps, from sensor to answer.

Three perspectives

Up close, from above and looking ahead.

Measured

Near the station

Temperature, pressure, CO₂, light, UV and soil, depending on the connected sensors. It relies on the physical hardware and connectivity.

Observed

From orbit

Satellite products for a wider area. A regional grid cell is not a measurement of one plant or one field.

Modelled

Looking ahead

Forecast maps for rain, temperature and wind. These are model outputs, not live weather radar.

NASA and Earth observation

No single product describes a whole farm.

So several are combined. Pick a source to see its orbit and how it should be read.

Landsat / HLS

NASA

705 km · 98.2° · 16-day repeat

What it is used for
Explore land reflectance and vegetation coverage.
How to read it
Cloud, shadow and water masks matter; a color image is not automatically an NDVI value.

Earthdata's CMR catalog discovers scenes and files; GIBS supplies the visual layers.

What users can do

A simple view, and a technical one for those who want more.

Clear summary

Forecast rain, temperature and wind alongside the local measurements available.

Forecast maps

Different times and a seven-day view where supported.

Satellite imagery

Historical imagery around the station, with a selectable date and area width.

Charts and history

Sensor series, plus access to NASA data from the technical view.

Ask the assistant

Free or suggested questions, for users who prefer not to type.

Missing data in plain sight

A missing reading is told apart from a valid measurement.

Scope and limitations

What the project does not claim.

  • 01

    Scientific point extraction is available on the deployed site. Some scenes may have no valid pixels because of cloud, quality, or coverage limitations.

  • 02

    Imagery is historical and may be obscured by clouds. Some products cover large regions or are published with a delay.

  • 03

    NASA fill values such as -999 are treated as missing data, not as a measurement.

  • 04

    Raw soil moisture needs calibration before it can be read as a percentage or used as an irrigation threshold.

  • 05

    The assistant summarizes what is available; it does not make unsupported diagnoses or automatic irrigation decisions.

Technology

From field hardware to the cloud.

Station and sensors

BMP390

Temperature and pressure

SCD30

CO₂, temperature and humidity

LTR390

Light and UV

Soil

Soil moisture and temperature

GPS

Coordinates associated with the station

Software and services

Languages

  • TypeScript
  • JavaScript
  • Python
  • SQL
  • HTML
  • CSS

Web application

  • Next.js 14
  • React 18
  • Node.js
  • Server Actions
  • API routes

Data

  • MySQL
  • mysql2
  • Zod

Interface

  • Tailwind CSS
  • Radix UI
  • Lucide
  • Recharts
  • Chart.js
  • dnd-kit
  • Three.js

Assistant

  • DeepSeek API
  • Vercel AI SDK

NASA

  • POWER
  • Earthdata CMR
  • GIBS

Imagery and maps

  • Planetary Computer
  • Sentinel-2
  • Google Maps
  • OpenTopoData

Forecasts

  • Meteomatics API

Local scientific processing

  • NumPy
  • h5py
  • netCDF4
  • Rasterio
  • pyproj
  • pyhdf

Development

  • Git
  • GitHub
  • ESLint
  • Node test runner
  • Vercel

The team

Made in Costa Rica.

A team that brings together software development, mechanical engineering and electromechanical engineering to connect field hardware with a more approachable way to explore agricultural data.

  • Software development
  • Mechanical engineering
  • Electromechanical engineering
  • Steven Alexander Alpizar Gamboa
  • Daniela Navarro Solís
  • Luis Mauricio Ureña Berrocal
  • Esteban José Retana Barboza
  • Sebastián Pastrana

Clearer information for decisions on the farm.

The aim is to make environmental information more approachable for the people who farm. No measured improvement in yield or water savings is claimed, because none has been validated yet.