NASA Space Apps Challenge 2026
KILLA.
Earth Change Observatory
An educational prototype for exploring environmental time series in Ecuador.
Team KILLA Esmeraldas
Esmeraldas, EcuadorGregorio Sebastian Gualavisi Gonzalez
Challenge: Be An Earth System Trend Detective!
AI-assisted code and writing · NASA data credited separately
The learning problem
Environmental claims need inspectable evidence
Teachers and students need a clear path from a regional question to a source, a trend estimate and its uncertainty.
KILLA makes that path visible in an interface that keeps the evidence close to the explanation.
Example investigation
Has the sampled area warmed over the available record?
The learner examines the variable, the period and the assumptions behind the answer.
The official challenge asks teams to visualize Earth science time series and quantify the size and significance of trends.
AI-assisted code and writing · NASA data credited separately
NASA data
NASA POWER, with source context
25 annual observations per series, covering 2001–2025.
NASA POWER supplies MERRA-2 reanalysis estimates for sample cells near Esmeraldas and Cuenca. The meteorological grid is 0.5° latitude by 0.625° longitude.
T2MAnnual mean air temperature at 2 metresDegrees Celsius (°C)
PRECTOTCORRAnnual mean daily precipitationMillimetres per day (mm/day). These values are not annual totals.
The source and processing record belong with every result.
Source: NASA POWER API v2.10.0, MERRA2. Monthly and Annual API · Meteorological sources. Annual YYYY13 entries retain the reported NASA values.
AI-assisted code and writing · NASA data credited separately
An example from the record
Temperature trends in the selected cells
| Sample cell | Change per decade | Nominal 95% interval |
|---|---|---|
| Near Esmeraldas0.96° N, 79.65° W | +0.329 °C | +0.119 to +0.539 °C |
| Near Cuenca2.90° S, 79.00° W | +0.132 °C | −0.118 to +0.382 °C |
These exploratory estimates describe the selected 25-year window and sample grid cells. They do not represent averages over an entire city or province. The intervals assume independent residuals and omit model and representation uncertainty.
NASA POWER annual T2M entries, 2001–2025. OLS calculations and full-precision results appear in the project's downloadable data record. Open the live data view.
AI-assisted code and writing · NASA data credited separately
Method and limits
Trend estimates need uncertainty
The fitted slope describes the estimated change per decade. The nominal 95% interval uses the Student t distribution with n − 2 degrees of freedom.
Ordinary least squares is an accessible starting point for investigating historical time series.
Exploratory interpretation
Climate records can contain correlated years. A basic OLS interval assumes independent residuals, so its coverage needs further checking.
Four location-variable comparisons remain exploratory, with no adjustment for multiple comparisons. The prototype omits p-values.
Causal attribution and local verification require further evidence.
AI-assisted code and writing · NASA data credited separately
Classroom use
An investigation students can explain
A teacher can turn one environmental question into a short exercise in reading evidence.
Learners compare locations and discuss how the selected variable, period and assumptions shape the answer.
Before inspecting the record
Write a claim and identify the observations needed to assess it.
After inspecting the record
Explain the trend, cite the source and name a limitation that affects the claim.
This is a proposed teaching activity. Classroom outcomes require evaluation.
AI-assisted code and writing · NASA data credited separately
Development and disclosure
A path toward a global observatory
First demonstration
Selected NASA POWER series, two Ecuadorian locations and transparent trend estimates for educational exploration.
Further development
Additional regions, deeper quality checks, classroom resources and validation with local expertise.
AI assistance: AI tools assisted with initial ideation, reference research, implementation and English writing. NASA provides the credited data products. The project makes no claim of operational forecasting or validated classroom impact.
References: NASA POWER · Official challenge
Team KILLA Esmeraldas, Ecuador
AI-assisted code and writing · NASA data credited separately