# oka.monitor Specification

:::info
Up to date by February 2023
:::

## Resources

- [Figma](https://www.figma.com/file/5kscCzjiDykrjl5dMWnrgq/processo?node-id=0%3A1)

## Definitions

- oka.monitor is an product for visualizing Historical and Live information on STMs (Spatio-Temporal Metrics) that are associated with Carbon Stocks generally.
- All STMs are computed for an given Reference Region that is provided by the User.

## STM Product Groups

#### Land Use Assessment Group

- Forest & Anthropic Land Cover Fraction
- Value range: $\{0, 1\}$ for each Cell.
- Spatial Resolution: 30m
- Temporal Resolution: 1 month with Landsat only. 2 weeks if including Sentinel.
- Spatial Aggregation: Average of Cells
- Methodology: Map Biomas v7.0 will be used as the base methodology [^mapbiomas_v7_atbd]. It is an comprehensive annual LCLU product that uses an variety of filtering and classification techonology that has been developed and validated by an consortium of Brazilian universities, NGOs, tech companies and public institutions. Specifically, it works by using an Random Forest & U-Net classifiers (p. 28) trained on top of independently trained samples. Transition and temporal filters are applied in order to categorize pixels on which the classifier was unable to provide an answer. Landsat is used for the Base methodology, although there's research on augmenting it with Sentinel 1 and 2 synergestically. [^mapbiomas_sentinel]
- Data Sources Required: Landsat 9
- Present Carbon Stock Estimate
- Value range: $\mathbb{R}_+$ for each Cell.
- Unit: tCO2
- Spatial Resolution: 30m
- Temporal Resolution: 1 month with Landsat only. 2 weeks if including Sentinel.
- Spatial Aggregation: Sum of Cells
- Temporal Resolution:
- Methodology: Estimating the Present Carbon Stock for Forests is typically done by starting off with an base year (eg. 2000) and using an Carbon Flux model for computing the periodical changes over time. We'll adopt for this STM the methodology laid out by Magliocca et al. [^carbon_stock], the same one adopted by the [Global Forest Watch project](https://www.globalforestwatch.org/map/) for Tree Biomass Density.
- Data Sources Required: Landsat
- Potential Carbon Stock Estimate
- Value range: $\mathbb{R}_+$ for each Cell.
- Unit: tCO2
- Spatial Resolution: 30m
- Temporal Resolution: 1 month with Landsat only. 2 weeks if including Sentinel.
- Spatial Aggregation: Sum of Cells
- Methodology: Computed by creating an Counterfactual Scenario on which all anthropized land cover is set to be Natural Forests. The Present Carbon Stock estimation methodology is then applied on this new scenario and the value is assigned for each grid cell. This is compatible with the Scenario-based Land Change Modeling Framework approached on the literature. [^counterfactual_lulc]
- Data Sources Required: Present Carbon Stock and Anthropic Land Cover Datasets

### Water Use Assessment

#### Water Quality Group

##### Water Clarity

- Predicted Secchi Disk Depth
- Value Range: $R_+$ for each cell
- Unit: meters
- Spatial Resolution: 30m
- Temporal Resolution: 1 month
- Spatial Aggregation: Average of Cells
- Methodology: An Random-Forest model (the best performing ML method according to Maciel et al.[^water_clarity]) will be trained using an reference dataset that has the raw reflectances, Normalized Difference Chlorophyll Index (NDCI) and Reflectance ratios as Features. The training dataset will be acquired from the Maciel et al.[^water_clarity] study.
- Accuracy: MSME=22%
- Data Sources Required: Sentinel-2 / MSI. Measurements:
- Reflectance data for the 490, 560, 660, 705 and 740 nm spectral bands

##### Trophic State

- Chlorophyll-a Concentration
- Value Range: $R_+$
- Unit: [Nephelometric Turbidity Unit](https://en.wikipedia.org/wiki/Nephelometer)
- Spatial Resolution: 30m
- Temporal Resolution: 1 month
- Spatial Aggregation: Average of Cells
- Accuracy: $MSME=28\%, R^2=0.80$
- Methodology: Sentinel-2 MSI TOA Level 1C reflectance images will be transformed through an empirical three-band spectral index. The parameters are to be calibrated through an sample dataset. [^chlorophyll]
- Data Sources Required: Sentinel-2/MSI

#### Water Balance Group

- Water Inflow
- Value Range:
- Spatial Resolution:
- Temporal Resolution:
- Spatial Aggregation
- Methodology:
- Data Sources Required:
- Water Outflow
- Value Range:
- Spatial Resolution:
- Temporal Resolution:
- Spatial Aggregation
- Methodology:
- Data Sources Required:

## Research Notes

- From [Evaluation of medium to low resolution satellite imagery for regional lake water quality assessments](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2011WR011005)
- "Two related water quality characteristics of inland lakes that are measurable by satellite imagery are trophic state and water clarity."
- "Trophic state typically is evaluated in terms of total phosphorus (TP) and chlorophyll a (chl a) concentrations and Secchi disk transparency (or Secchi depth)"
- [Trophic state](h[ttps://](https://en.wikipedia.org/wiki/Trophic_state_index)): "The Trophic State Index (TSI) is a classification system designed to rate water bodies based on the amount of biological productivity they sustain"
- "Water clarity variables include turbidity, suspended solids, and Secchi depth."
- [Water clarity](https://en.wikipedia.org/wiki/Water_clarity): "Water clarity is a descriptive term for how deeply visible light penetrates through water. In addition to light penetration, the term water clarity is also often used to describe underwater visibility."
- "Except for TP, these variables have optical properties that can be inferred from multispectral satellite imagery"
- **Mental note**: Those indexes are about Water Quality as a topic. The link to Water Use as a topic would likely be as a Effect given the Causes. One lead then is to identify Use-Related Assessments
- From https://en.wikipedia.org/wiki/Water_quality
- Water quality refers to the chemical, physical, and biological characteristics of water based on the standards of its usage
- The complexity of water quality as a subject is reflected in the many types of measurements of water quality indicators
-

## References

[^mapbiomas_v7_atbd]: [MapBiomas General “Handbook”, Algorithm Theoretical Basis Document (ATBD), Collection 7, Version 1.0](https://mapbiomas-br-site.s3.amazonaws.com/ATBD_Collection_7_v2.pdf)
[^mapbiomas_sentinel]: [Brandmeier, M., Hell, M., Cherif, E., and Nüchter, A., “Synergetic use of Sentinel-1 and Sentinel-2 data for large-scale Land Use/Land Cover Mapping”, 2022. doi:10.5194/egusphere-egu22-4300.](https://ui.adsabs.harvard.edu/abs/2022EGUGA..24.4300B/exportcitation)
[^counterfactual_lulc]: [Nicholas R. Magliocca, Pratik Dhungana & Carter D. Sink (2023) Review of counterfactual land change modeling for causal inference in land system science, Journal of Land Use Science, 18:1, 1-24, DOI: 10.1080/1747423X.2023.2173325](https://www.tandfonline.com/doi/full/10.1080/1747423X.2023.2173325?scroll=top&needAccess=true&role=tab)
[^carbon_stock]: Harris, N.L., D.A. Gibbs, A. Baccini, R.A. Birdsey, S. de Bruin, M. Farina, L. Fatoyinbo, M.C. Hansen, M. Herold, R.A. Houghton, P.V. Potapov, D. Requena Suarez, R.M. Roman-Cuesta, S.S. Saatchi, C.M. Slay, S.A. Turubanova, A. Tyukavina. 2021. Global maps of twenty-first century forest carbon fluxes. Nature Climate Change. https://doi.org/10.1038/s41558-020-00976-6

[^water_clarity]: Daniel Andrade Maciel, Claudio Clemente Faria Barbosa, Evlyn Márcia Leão de Moraes Novo, Rogério Flores Júnior, Felipe Nincao Begliomini,
Water clarity in Brazilian water assessed using Sentinel-2 and machine learning methods, ISPRS Journal of Photogrammetry and Remote Sensing, Volume 182, 2021, Pages 134-152, ISSN 0924-2716, https://doi.org/10.1016/j.isprsjprs.2021.10.009.

[^chlorophyll]: Aranha, T.R.B.T.; Martinez, J.-M.; Souza, E.P.; Barros, M.U.G.; Martins, E.S.P.R. Remote Analysis of the Chlorophyll-a Concentration Using Sentinel-2 MSI Images in a Semiarid Environment in Northeastern Brazil. Water 2022, 14, 451. https://doi.org/10.3390/w14030451