Open assessment and Opasnet Jouni Tuomisto 16062021 1

Open assessment and Opasnet Jouni Tuomisto 16/06/2021 1

Outline / Menu • • • Motivation Examples Assessment structure Technical functionalities Challenges Way forward 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 2

An impact assessment in Opasnet 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 3

Climate change policy model 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 4

Links to the pages of climate change policy model • http: //en. opasnet. org/w/Climate_change_policies_an d_health_in_Kuopio • http: //en. opasnet. org/w/Building_model • http: //en. opasnet. org/w/Building_stock_in_Kuopio • http: //en. opasnet. org/w/Emission_factors_for_burnin g_processes • http: //en. opasnet. org/w/Exposure_to_PM 2. 5_in_Finl and • http: //en. opasnet. org/w/Population_of_Kuopio 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 5

Motivation • Open assessment aims to develop impact assessment practices and improve the information flow between experts and decision makers. • Research question: How can scientific information and value judgements be organised to support decision making in such a way that open participation is allowed? 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 6

Basic structure of a page 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 7

Six principles are applied: – Intentionality: All that is done aims to offer better understanding to the decision maker about outcomes of the decision. – Shared information objects: all information is shared using a systematic structure and a common workspace where all participants can work. – Causality: The focus is on understanding the causal relations between the decision options and the intended outcomes. – Critique: All information presented can be criticised based on relevance and accordance to observations. – Reuse: All information is produced in a format that can easily be used for other purposes by other people. – Openness: All work and all information is openly available to anyone interested. Participation is free. If there are exceptions, these must be publicly justified. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 8

Opasnet web-workspace • Motto: You must be able to do any work in Opasnet, whether you are a researcher or a decision-maker. • It is designed for effective information management and use. • Parts: – – 16/06/2021 Opasnet wiki: the user interface R statistical software for online models Mongo. DB database for all kinds of data File management system. Marjo Niittynen, Jouni T. Tuomisto 9

Challenges • Surprisingly: lack of interest in participating. • Openness seen as threat or potential cause of trouble. • Researchers are not familiar with R. • Decision makers have no time nor resources to do anything extra. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 10

Why Opasnet and ovariables? • The six principles are strict, and most existing data sharing or assessment practices violate several of them. • The ovariable structure and the databases in Opasnet arguably offer a method to fulfil all principles at the same time. • Because they are very generic, they could be used as data management, storage, and archive standards. But only if – people know them well enough to routinely use them, – coworkers use them as well. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 11

Threshold of new practices • Changing working practices is always difficult, because you know your current practices. • In the case of ovariables, most benefits only come when many other people use them as well. – To a primary researcher the benefits are not obvious, because they mostly come in impact assessment. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 12

Take-home messages • Open assessment is a method in production. • Opasnet can be used by anyone. • A critical mass of experts / modellers is still missing. – When we get that, the system becomes increasingly more effective and attractive. • NOW is the time to join Open Assessors’ Network. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 13

Basic idea of ovariables: example • How to handle complex information in a model in a standardised and efficient way? • Three grams of SO 2 deposits onto the ground per m 2. What is the total deposition on the area of 2 m 2? • 3 g /m 2 * 20 m 2 = 60 g • Simple to do with any software. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 14

More complex example • The deposition is four and three grams per m 2 of SO 2 in 2011 and 2012, respectively. What is the total deposition on the areas of 20 and 100 m 2? • On a spreadsheet, you could put years as rows and areas as columns. However, if you have to add a third dimension, your approach falls apart. • Instead, you could do it a bit more cleverly: 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 15

How to use 2 D tables for multidimensional data Area Year Deposition rate Total deposition 20 2011 4 80 20 2012 3 60 100 2011 4 400 100 2012 3 300 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 16

The idea of information objects • Object area (units m 2) Place Result Patio 20 Garden 100 • Object deposition. rate (units g /m 2) Year Result 2011 4 2012 3 – Explanation column (index) Observation column (result) • The usability is different than that of a spreadsheet. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 17

Calculations with ovariables • total. deposition <- deposition. rate * area Place Year deposition. ra area. Result te. Result total. depositi on. Result Patio 2011 4 20 80 Patio 2012 3 20 60 Garden 2011 4 100 400 Garden 2012 3 100 300 • Indices are used to combine matching rows • Intermediate results are shown as well. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 18

How many rows in the output with these indices? Place Year SO 2 policy Pollutant Patio 2011 BAU SO 2 Garden 2012 Active NOx 2013 2014 • 32 rows – at most. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 19

How ovariables are merged Year SO 2 policy Pollutant policy. Result 2011 BAU SO 2 1 2011 Active SO 2 0. 9 2013 BAU SO 2 1 2013 Active SO 2 0. 7 Year Pollutant deposition. r ate. Result 2011 SO 2 4 2012 SO 2 3 2013 SO 2 2. 8 2011 NOx 1 2012 NOx 0. 9 2013 NOx 0. 9 16/06/2021 Place area. Result Patio 20 Garden 100 • In this case, only 8 rows merge successfully. Marjo Niittynen, Jouni T. Tuomisto 20

Data storage in Opasnet • Opasnet. Base has several interfaces – Table 2 Base <t 2 b> – Opasnet. Base. Import (Upload data and Show results) – opbase. upload() and opbase. data() • Opasnet server for R objects – objects. put() and objects. get() – objects. store() and objects. latest() • Opasnet wiki for files – Upload file and opasnet. data, opasnet. csv • M-files for permanent documents, eg. Reports 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 21

Opasnet. Base • Versatile database for 2 D tables. • Data can be uploaded from – User interface Opasnet. Base. Import – Wiki page using <t 2 b> – From R using opbase functions 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 22

Opasnet server • Stores model runs and R objects created in models. • Is used to store intermediate objects in models. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 23

Opasnet wiki for files • Upload file functionality in wiki • Files can be accessed from R using functions opasnet. data and opasnet. csv 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 24

Building stock, climate policies and health in cities Jouni Tuomisto, Marjo Niittynen, THL Marjo Niittynen, Jouni T. Tuomisto 16/06/2021 25

Background • Buildings are in a key role when reducing greenhouse gas (GHG) emissions. – They consume energy in heating and cooling. – Different heating solutions differ in emissions and health impacts. • Buildings affect health in several ways. – Ventilation, heat/cold, dampness, biomass burning, radon, smoking, space, neighbourhood etc… • How to reduce GHG emissions and promote health at the same time? 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 26

Buildings in Basel by floor area 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 27

Buildings by age on map 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 28

Building stock model 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 29

Building floor area in Basel 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 30

Building energy need in Basel 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 31

Emissions in Basel 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 32

CHP plant in Haapaniemi 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 34

The study • Building stock model to predict past and future energy need of Kuopio. • Includes two policies: – Fuel policy • BAU: in Haapaniemi CHP plant, 85 % of peat is used. • Biofuel increase: 25 % peat, 60% wood. – Renovation policy • BAU: 3 % of buildings is energy-renovated per year • Active renovation: 4. 5 % per year • Effective renovation: more amibitious renovations 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 35

Causal chain • • • 16/06/2021 Building stock (floor area) Heating energy need Heating systems PM 2. 5 and CO 2 emissions Dispersion of and exposure to PM 2. 5 Health impacts of PM 2. 5 Marjo Niittynen, Jouni T. Tuomisto 36

Building stock 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 37

Heating energy 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 38

Emissions 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 39

Health impacts 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 40

Conclusions from the case studies • Health impacts of district heating are small in all Kuopio scenarios (in the order of 10 DALY / year) • However, the impact of primary wood heating in houses is relatively large (7 DALY / year) • Assumption of zero-carbon-emission from wood should be revisited using the LCA. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 49

Conclusions from the building model • An online building stock model was developed and connected to energy use and health impact models. – The model is robust, i. e. it works with data with different level of coarseness. – It will be combined with the energy balance model before the end of project. • Availability (or sharability) of data is a major issue – Model should be usable by cities themselves. • More discussion is needed between policy makers and researchers/experts. – Development of practical scenarios. – Providing scientific info in readily-usable format. 16/06/2021 Marjo Niittynen, Jouni T. Tuomisto 50