NASA Space Apps 2026 · Field Shift
One index for climate, water, soil and food
Fihrist reads 24 years of NASA data for any field and compares three 3-year crop rotations: the one a farmer uses today and two adaptive ones. Water need, soil nitrogen, calories and protein per hectare, and the risk across thousands of possible futures. Every number is traceable to a dataset or a cited source.
The problem
Farm decisions need many datasets. Tools look at one at a time.
In Tajikistan and northern Afghanistan, summer heat passes 40 °C, irrigation water is shared and limited, and years of continuous wheat or cotton draw down soil nitrogen. Climate, water, soil and nutrition are usually analysed in separate tools. A farmer choosing next year's crop has to combine them alone.
Water
How much irrigation does each rotation need in a normal year, a dry year, a hotter future?
Soil
Which rotation rebuilds nitrogen with legumes, and which one runs it down and needs more fertiliser?
Food and income
How many people can one hectare feed, and where does the familiar rotation still earn more in year one?
How it works
From a field outline to three explained rotations
The equations produce the numbers; the interface only explains them. The method is open source and covered by unit tests, including FAO-56 worked examples.
Pick a field
Fly to it on a 3D globe and outline the field, or start from one of four preset regions.
Read NASA data
Daily NASA POWER weather (2001–2024), MODIS crop greenness, root-zone soil wetness and NASADEM terrain.
Run open equations
FAO-56 Penman–Monteith evapotranspiration, a daily soil-water balance, FAO-33 yield response and a nitrogen balance.
Compare and decide
Three rotations, each with why it is suggested, NASA evidence, benefits, data gaps and local risks for the farmer to review.
What only the combination shows
Where datasets disagree, there is a question worth asking
Fihrist puts weather, soil water and satellite greenness on one monthly grid and compares each month with the field's own 24-year history. It flags months where crops were much less green than usual although heat, rain and soil water were normal. The cause lies outside the weather record.
First result: all three Central Asian presets show this in March–April 2012; Kansas does not. The winter before was unusually cold at all three sites. Winter-kill or late green-up would fit, but this is a hypothesis to confirm with farmers, not a finding. The three sites share one regional winter.
| Winter 2011–12 | Gissar | Khatlon | Balkh |
|---|---|---|---|
| Frost days, February | 27 (19.1) | 24 (10.9) | 21 (9.6) |
| Frost days, March | 19 (7.5) | 11 (3.0) | 8 (2.5) |
| Crop NDVI, March 2012 | 0.26 (0.42) | 0.19 (0.28) | 0.17 (0.28) |
Normal (2001–2024 mean) in brackets. NASA POWER daily minimum temperature; MODIS Terra NDVI averaged over ESA WorldCover cropland within 5 km.
Validation
Validated in Kansas, where ground truth exists
Rainfed winter wheat simulated at the Kansas preset, compared with USDA NASS statewide yields for 2015–2024. We report the gap instead of tuning it away.
Year-to-year correlation with real Kansas yields
Years in which the model moved in the same direction as NASS
Mean level too low (1.64 vs 2.99 t/ha), so rainfed numbers compare rotations; they do not forecast a harvest
Who it's for
Built for the farmer, auditable by the scientist
What works today, and what comes next.
Farmers and extension agents
Outline a field, set the rotation you use today, choose your priorities (water, nitrogen, income and more) and compare three plans in your language.
- Next: monthly risk alerts by SMS for farmers without internet
Scientists
Every equation, constant and source is shown in the app and in the open repository, with the validation results and the methods we rejected.
- Next: CSV export and an open API
- Next: checks against flux towers and field cameras
Cooperatives, NGOs, companies
Many fields in one view: which farms face compound stress next season, and which rotations lower water and fertiliser costs.
- Accounts and saved fields
- Gross margin per hectare with local costs
- Export to Excel and finance tools
About
Why "Fihrist"
Fihrist means "index" in Persian and Arabic. In the 10th century, Ibn al-Nadim's Kitab al-Fihrist catalogued the knowledge of his time across disciplines. Scholars of the same tradition, such as Al-Biruni and Ibn Sina, worked across many fields at once. Fihrist follows that idea: connect separate datasets and disciplines in one auditable index.
Fihrist was built for the 2026 NASA Space Apps Challenge, challenge Field Shift: Adapting Farms with NASA Data. The code is open source under the MIT license.
Beyond Earth
The same limits, further away
Every permanent settlement beyond Earth will have to farm in a closed loop. Water-stressed farms on Earth already live near those limits. The measures Fihrist reports (calories per cubic metre of water, nitrogen balance, people fed per hectare) are the same kinds used to evaluate bioregenerative life support. Fihrist shares those constraints; it does not design space habitats.
See your field through 24 years of NASA data
Free and open source. Works on a phone.