early version

_build/html/.buildinfo

@@ -0,0 +1,4 @@
+# Sphinx build info version 1
+# This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done.
+config: 041481c9bbd721076049b1dc858ca2f1
+tags: 645f666f9bcd5a90fca523b33c5a78b7

_build/html/_sources/project.md

@@ -0,0 +1,112 @@
+# Proposal
+
+The challenge in front of us is to transform the observations into accurate and actionable information for two different purposes. There is the 
+administration of the territory and identification of priorities with regard to plastics in the environment. Then, once the priorities have been established these 
+new indicators must be reported on. This project concerns the infrastructure and processes necessary to identify priorities and develop indicators to reduce plastics 
+in the environment from litter density counts.
+
+Therefore, we propose a harmonised reporting system that is administered by regional or national NGOS in collaboration with an academic partner with 
+the stated goal of producing reliable reports for the public administration and others who communicate on such matters. This allows administrations to 
+identify priorities and set baseline values according to agreed upon and observed indicators. 
+
+The first report proved that a relatively small team could cover a large territory with the correct technology. This remains the case. 
+That is why we see this project, in its current iteration, specifically attractive to stakeholders with an internal data-science team. 
+Stakeholders with experience in python or R will feel right at home in the development environment and will be the source of many improvements. 
+For those interested in learning the basics, we are here to coach them and help them in their projects.
+
+The proposal is thus to establish a hub that is dedicated to the collection, cleaning, aggregating and distribution of litter density data. 
+The distribution of the data has the primary goal of producing decision support products to local administrations and other organizations charged with 
+reducing plastics in the environment. The _hub_ is staffed by three people and who conduct surveys, produce value added research, teach other 
+NGOs to collect samples, and respond to requests from local administrations. Activities or changes in protocol are considered in consultation with academic advising team. 
+
+## Goals
+
+> * Develop a reporting and research tool for stakeholders in the waste management, life cycle assessment and environmental monitoring sectors.
+> * Increase capacity of stakeholders to monitor and report observations according to the current standard in the federal report using open source tools.
+> * Assist in the reduction and prevention of plastics in the environment
+
+:::{dropdown} Deliverables: four year contract
+1. 800 samples
+2. Revised updated federal report
+3. New cantonal and municipal reports
+4. Cloud service for cantonal and municipal reporting
+5. Chat service for endusers to develop indi
+:::
+
+:::{dropdown} Objectives
+1. Collect 400 - 500 samples in a 12 month period
+   * reduce the number of items reported as _fragmented_ or _Gfrags_ 
+     1. This value can range from 20% - 30% of the total items collected
+   * include polymer identification for a subset of samples
+     1. Subset should support an academic publication
+     2. Subset should support a regional assessment of polymer types
+   * Increase knowledge sharing between surveyors
+   * Capacity building for surveyors
+
+2. Increase the number of associations competent to collect and report data
+   * Develop one association on Lac de Constance or Zurichsee
+   * Develop training materials and protocols with ASL 
+
+3. Structure report specifically for cantonal and municipal reporting
+   * Revise and improve the current report
+   * Consider the format of the report given cantonal requirements (attached) 
+
+4. Build upon previous research and reports
+   * Continue to develop modeling and parameter estimation for litter density
+   * Focus on topographical features and densities
+   * Consider polymer types and densities
+
+5. Define groups of objects of interest with stakeholders
+   * single use plastics
+   * agricultural plastics
+   * construction plastics
+   * industrial plastics
+
+6. Develop a cloud first infrastructure and chat agent
+   * Move all data and software assets to Infrastructure as a Service (IaaS)
+   * Develop machine learning and artificial intelligence pipelines for data analysis
+   * Develop Automated programming interfaces (APIs) for data collection and reporting
+   * Develop a user interface for data collection and reporting
+   * Deploy chat agent
+:::
+
+:::{dropdown} Benchmarks
+1. Year one: Proposed updates and revisions to the federal report. Proposed model for cantonal and municipal reporting.
+   * Formal research proposal for polymer identification
+   * Formal research proposal for topographical features
+   * Functioning prototype of cloud service
+   * 100 samples collected
+   * stakeholder meeting
+2. Year two: 
+   * Assessment of software and data collection
+   * Revised report with current data
+   * Update on capacity building
+   * End of year two:
+     1. Presentation of revised report with current data
+     2. Summary of capacity building
+     3. 400 samples
+     3. stakeholder meeting
+3. Year three:
+   * Presentation of feedback
+   * stakeholder meeting
+   * Proposed changes to app and report
+   * Final report:
+     1. implementation of feedback
+     2. 200 samples
+     3. cloud service for cantonal and municipal reporting
+4. Year four:
+   * Testing and implementation
+   * 100 samples
+
+:::
+
+:::{dropdown} Cost
+
+> Less than 1% of the estimated cost to prevent and remove litter annualy in Switzerland. [BAFU/OFEV - litter costs](https://www.bafu.admin.ch/bafu/en/home/topics/waste/publications-studies/publications/litter-dropping-costs-money.html)
+
+:::
+
+
+## Interested ?
+
+Interested administrations or other potential financial partners should contact roger at hammerdirt.

_build/html/_sources/references.md

@@ -0,0 +1,74 @@
+# References
+
+## Collection protocols and previous studies in Switzerland
+
+(mlw_2013)=
+__Guidance on Monitoring of Marine Litter in European Seas__, Galgani et all,  EUR 26113. Luxembourg (Luxembourg): 
+Publications Office of the European Union; 2013. JRC83985 [https://publications.jrc.ec.europa.eu/repository/handle/JRC83985](https://publications.jrc.ec.europa.eu/repository/handle/JRC83985)
+
+(mlw_2023)=
+__Guidance on the Monitoring of Marine Litter in European Seas An update to improve the harmonised monitoring of 
+marine litter under the Marine Strategy Framework Directive__,  Galgani et all., EUR 31539 EN, Publications Office of the European Union, Luxembourg, 2023, 
+ISBN 978-92- 68-04093-5, doi:10.2760/59137, JRC133594 [https://literatur.thuenen.de/digbib_extern/dn067364.pdf](https://literatur.thuenen.de/digbib_extern/dn067364.pdf)
+
+(eu_threshold)=
+__A European threshold value and assessment method for macro litter on coastlines__, Publications Office of the 
+European Union, Van Loon, et all , [https://publications.jrc.ec.europa.eu/repository/handle/JRC121707](https://publications.jrc.ec.europa.eu/repository/handle/JRC121707), ISBN 978-92-76-21444-1, doi:10.2760/54369, JRC121707.
+
+(basin_organization)=
+__Transfer of waste and plastics in aquatic environments__, Christophe Brachet, International Netwwork of Basin Organizations, 2024, [INBO](https://www.inbo-news.org/)
+
+(land_use_role)=
+__Revealing the role of land-use features on macrolitter distribution in Swiss freshwaters__, Schreyers et al., 2022, Environmental Pollution Journal, [https://doi.org/10.1016/j.envpol.2024.124911](https://doi.org/10.1016/j.envpol.2024.124911) article: [sciencedirect](https://www.sciencedirect.com/science/article/pii/S0269749124016257)
+
+(ofev_dechets_cours_deau)=
+__Des déchets aux abords des lacs et cours d’eau suisses__, Office fédéral de l'environment, 2021, [OFEV](https://www.bafu.admin.ch/bafu/fr/home/themes/dechets/dossiers/des-dechets-aux-abords-des-lacs-et-cours-d-eau-suisses.html)
+
+(plastock)=
+__Plastock: Estimation of plastics on the shores of Lake Geneva, switzerland__, _Pochelon et al._, 2023, [Plastock](https://associationsauvegardeleman.github.io/plastock/)
+
+(OFEV)=
+__Identification, quantification and analysis of observable anthropogenic litter along Swiss lake systems__, Erismann et al.,  OFEV. 2021. [IQAASL](https://hammerdirt-analyst.github.io/IQAASL-End-0f-Sampling-2021/index.html)
+
+(SLR)=
+__Swiss Litter Report__, Gabrielle, Kuhl; Pascal, Blahrer; 2018; [Stopp research](http://stoppp.org/)
+
+## Math and geospatial
+
+(scipy)=
+__scipy: Fundamental Algorithms for Scientific Computing in Python__, _Virtanen et al_, Nature Methods, 2020, doi: 10.1038/s41592-019-0686-2, [scipy](https://www.scipy.org/).
+
+(pandas)=
+__pandas__, _The pandas development team_, Python software, 2020, doi: 10.5281/zenodo.3509134, url: [https://doi.org/10.5281/zenodo.3509134](https://doi.org/10.5281/zenodo.3509134).
+
+(python_3)=
+__Python 3 Reference Manual__, _Guido Van Rossum and Fred L. Drake_, CreateSpace, 2009, isbn: 1441412697, [python](https://www.python.org/).
+
+(geopandas)=
+__geopandas__, _Joris Van den Bossche et al_, Python software, 2020, doi: 10.5281/zenodo.3509134, url: [https://doi.org/10.5281/zenodo.3509134](https://doi.org/10.5281/zenodo.3509134).
+
+(swisstlm)=
+__swissTLMRegio: Le modèle numérique de la Suisse à petite échelle__, Swiss Federal office of topography, [swiss topo](https://www.swisstopo.admin.ch/fr/modele-du-territoire-swisstlmregio)
+
+(myst)=
+__MyST: Markedly Structured Text__, _Executable Book Project_, 2021, [https://myst-parser.readthedocs.io/en/latest/](https://myst-parser.readthedocs.io/en/latest/)
+
+(scikit-learn)=
+__scikit-learn: Machine Learning in python__, _Pedregosa et al_, Journal of Machine Learning Research, 2011, doi: 10.1109/MCSE.2007.55, [scikit-learn](https://scikit-learn.org/stable/).
+
+(matplotlib)=
+__Matplotlib: A 2D graphics environment__, _Hunter, J. D._, Computing in Science & Engineering, 2007, doi: 10.1109/MCSE.2007.55, [matplotlib](https://matplotlib.org/).
+
+(bayes_gelman)=
+__Bayesian Data Analysis - Third edition__, _Andrew Gelman_, Columbia university, 2021, [http://www.stat.columbia.edu/~gelman/book/BDA3.pdf](http://www.stat.columbia.edu/~gelman/book/BDA3.pdf), doi: 10.1201/b16018.
+
+(think_bayes)=
+__Think Bayes 2ed__, _Alan Downey_, Green Tea Press, 2021, [http://allendowney.github.io/ThinkBayes2/](http://allendowney.github.io/ThinkBayes2/)
+
+(bayes_hackers)=
+__Bayesian Methods for Hackers__, _Cameron Davidson-Pilon_, Addison-Wesley Professional, 2021, [https://github.
+com/CamDavidsonPilon/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers](https://github.com/CamDavidsonPilon/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers),
+doi: 10.5555/2851115
+
+(bayes_jaynes)=
+__Probability Theory: The logic of science__, _Edwin Jaynes_, Washington University, 2003, isbn: 0521592712. [the logic of science](https://bayes.wustl.edu/etj/prob/book.pdf)

_build/html/_sources/the_llm_assistant.md

@@ -0,0 +1,58 @@
+# The role of the llm assistant
+
+__Definition of the situation, problem and solution__
+
+## Reporting volunteer observations
+
+Alot of ink has been spilled on the topic of citizen science and the role of volunteers in data collection. We assume that a 
+a reasonable person would place value on the collective observations of a segment of the population in reference to a particular
+topic. We also assume that the value of any such collective observations is largely dependent on the method in which the 
+observations were made and the transparency of the analytical process.
+
+Beach litter data has two broad purposes: 1. To inform policy and 2. To inform the public. However, an informed policy at 
+the national level is not the same at the regional level. Resource allocation and policy implementation are at the 
+cantonal and municipal level. Thus, reasonable people at various levels of the administrative hierarchy may not have access to
+the information needed to make informed decisions.
+
+## Beach litter surveys
+
+The European Environment Agency (EEA) has been working with volunteers to collect data on marine litter. The volunteers 
+have been collecting data on the amount and type of litter found on beaches in Europe. The collection method is defined
+in the _Guide to monitoring marine litter in European seas_. Based on the analysis of the collected data a suggested 
+threshold value, and a method to compute it was introduced in 2020. The threshold value is based on the principal of 
+precaution regarding exposure to plastics. In switzerland monitoring of shoreline litter along lakes and rivers
+has followed the same protocol with minor modifications due to topography since 2015. 
+
+## Data availability and reporting
+
+The availability of the data is largely dependent on local NGOs and volunteers. There is no legal requirement to report litter data.
+The federal report, modeled after the EEA report, provides a format and a method for aggregating and reporting but no 
+infrastructure to collect and report the data. This infrastructure is non-existent at the cantonal level. In the Léman region
+data collection and reporting has standardized because of the work of local NGOs. The most recent report was published in 2024. 
+This report was not aggregated at the cantonal or municipal level either.
+
+## Environmental monitoring and protection
+
+Environmental monitoring and protection are the responsibility of the cantons in Switzerland. The cantons can and do delegate
+certain responsibilities to the municipalities. Thus, a report at the national level provides an overview of the situation 
+but provides very little value in the way enabling action at the cantonal or municipal level. Even though the data is collected at the
+municipal level it is not aggregated and reported at the cantonal level.
+
+## Cantonal and municipal resources
+
+Each canton and municipality has a section or department responsible for environmental pollution. These departments, for the most part,
+are already working with limited resources. Managers of these departments may be more than competent in interpreting a report but lack the
+additional human resources necessary to collect and prepare reports for their own canton or municipality. The cause of this is two fold: 1. 
+The lack of common infrastructure for data collection and reporting and 2. Inexperience in domain specific data handling and analysis
+
+## The solution
+
+We provide a solution that maximizes the efficiency of the reporting process by leveraging large language models (llms) and 
+standardized scientific computing. The llm assistant produces a rough draft of the standardized report, ready to be reviewed and 
+edited by the domain expert (or the responsible person). This reduces the cost to report the data and keeps a human in 
+the analytical loop. More of the resources can be allocated to collection and communication of the data. Or if you look at this way, 
+small investments in data collection will lead to large returns in communication at the municipal and cantonal level.
+
+Ultimately, the goal is that those who must act to reduce pollution have the data in the format that they need to make informed 
+decisions. The llm assistant  uses the current monitoring methods and the capabilities of the llms and scientific
+computing to provide a solution that is both efficient, transparent and appropriate to all scales of governance and stakeholders.

_build/html/_sources/titlepage.md

@@ -0,0 +1,70 @@
+# Reporting volunteer observations
+
+This is the proof of concept for a distributed reporting system of marine litter density using data from volunteers.
+
+:::{dropdown} Recent changes
+
+__September 25:__ Added llm assisted reporting
+
+__June 24:__ Added responses to use cases
+
+__June 22:__ Added methods to calculate landuse profile using Geopandas.
+
+
+:::
+
+::::{grid}
+
+:::{grid-item}
+
+This is an application of large language models (llms) and machine learning to provide summary reports of volunteer observations
+of marine litter using the OSPAR, JRC or NOAA methods for counting beach litter. 
+
+The application provides summary reports of exising data and provides forecasts for a region, city or body of water. The  llm 
+acts first as a transcriber of test results and then as an assistant to provide a written report and resource package that can
+be either used directly or integrated into other documents. 
+
+:::
+
+:::{grid-item-card}
+
+```{image} resources/hammerdirt.png
+:alt: bob
+:class: bg-primary mb-1
+:width: 400px
+:align: center
+```
+:::
+::::
+::::{grid}
+
+:::{grid-item}
+
+
+This application minimizes the cost of reporting and evaluating beach litter density and uses geospatial analysis to leverage
+the use of data from other disparate locations. Most importantly, decision makers and stakeholders have access to reliable, 
+readable reports that allow them to communicate effectively and make informed decisions. 
+
+The data in these reports is a combination of observations from variety of groups in Switerland since 2015. The observations were recorded using an interpretation of the _Guide for Monitoring Marine Litter on European Seas_ [The guide](https://mcc.jrc.ec.europa.eu/main/dev.py?N=41&O=439&titre_chap=TG%20Litter&titre_page=Guidance%20for%20the%20Monitoring%20of%20Marine%20Litter). The guide and the monitoring of beach litter are part of decades of research, here is the brief history [A Brief History of Marine Litter Research](https://link.springer.com/chapter/10.1007/978-3-319-16510-3_1).
+
+:::
+:::{grid-item}
+__Associated projects__
+
+* https://hammerdirt-analyst.github.io/IQAASL-End-0f-Sampling-2021/
+* https://hammerdirt-analyst.github.io/solid-waste-team/titlepage.html
+* https://hammerdirt-analyst.github.io/landuse/titlepage.html
+* https://associationsauvegardeleman.github.io/plastock/
+* https://hammerdirt-analyst.github.io/finding-one-object/titlepage.html
+
+
+:::
+
+:::{grid-item}
+:columns: 12
+
+__THE TEAM__
+
+Shannon Erismann, Montserrat Filella, Téo Gursoy, Christian Ludwig, Roger Erismann, Bettina Siegenthaler
+:::
+::::