Rainbow color map distorts and misleads research in hydrology – guidance for better visualizations and science communication

Nowadays color in scientific visualizations is standard and extensively used to group, highlight or delineate different parts of data in visualizations. The rainbow color map (also known as jet color map) is famous for its appealing use of the full visual spectrum with impressive changes in chroma and luminance. Besides attracting attention, science has for decades criticized the rainbow color map for its non-linear and erratic change of hue and luminance along the data variation. The missed uniformity causes a misrepresentation of data values and flaws in science communication. The rainbow color map is scientifically incorrect and hardly decodable for a considerable number of people due to color vision deficiency (CVD) or other vision impairments. Here we aim to raise awareness of how widely used the rainbow color map still is in hydrology. To this end, we perform a paper survey scanning for color issues in around 1000 scientific publications in three different journals including papers published between 2005 and 2020. In this survey, depending on the journal, 16 %–24 % of the publications have a rainbow color map and around the same ratio of papers (18 %–29 %) uses red–green elements often in a way that color is the only possibility to decode the visualized groups of data. Given these shares, there is a 99.6 % chance to pick at least one visual problematic publication in 10 randomly chosen papers from our survey. To overcome the use of the rainbow color maps in science, we propose some tools and techniques focusing on improvement of typical visualization types in hydrological science. We give guidance on how to avoid, improve and trust color in a proper and scientific way. Finally, we outline an approach how the rainbow color map flaws should be communicated across different status groups in science.

Misinterpretation of hydrological studies in the Lancang-Mekong Basin: drivers, solutions and implications for research dialogue

Socio-hydrology presents one of the scientific approaches interpreting complex interactions between human and water systems. To date, water becomes extremely politicized by non-scientists and frequently put in a broader political context with non-water issues. The purpose of this text is to (1) analyse drivers of the growing politicization of hydrological science in the Lancang-Mekong Basin, (2) examine solutions for addressing the misinterpretation of hydrological data, and (3) outline the unintended consequences of politicization the hydrological studies. The paper argues that politicization of science (i) gives more power to non-scientists, (ii) undermines the trust in science and other research institutions, (iii) creates inequality among hydrological studies and water scientists, and (iv) provides more incentives for making research tailored to desirable outcomes. The topic is highly actual and beneficial for water experts and other scientists who want to better understand the potential negative implications of hydrological studies and the limits of socio-hydrology.

Losing faith

<p><span>The presented session examines the politicization of hydrological science and discusses the current implications for misinterpreting the hydrological data that undermine trust in science. As a result of growing medialization of hydrological studies and simplifying the research conclusions for the wide public, it is more difficult for hydrologists to keep scientific integrity and not fall into the realm of subjectivism. By close analysis of two hydrological studies (Pöyry Report and Eyes on Earth Studies), we noticed that (1) research conclusions may be tailored to political demand, (2) intentionally overlook basic theoretical-methodological research standards, and (3) negatively influenced by social media, especially when the research conclusions do not correspond with scientific reviews nor official speech acts from state authorities. On the other hand, we also found several unintended consequences that make politicization science useful and even positive, especially in terms of changing the social perception of water or deepening the water cooperation in hydrological monitoring which still remain sensitive political issues in many corners of the world.<br><br>By drawing on the socio-hydrology and critical hydropolitical theories, this session (i) explores the current challenges for interpreting the hydrological studies, (ii) clarify the techniques how to prevent misinterpretation of the hydrological data, and (iii) demonstrate the politicization of the hydrological science on two micro-case studies within the Mekong River Basin that raise controversies among scientists and potential disputes among states. While the Pöyry Report conclusions served as a political tool to justify the construction of Xayaburi hydropower dam in Laos regardless the opposition of downstream countries, the Eyes on Earth Study was designed to undermine mutual trust among Mekong states and damage the credibility of other hydrological studies that do not share the same opinion on hydrological changes in the Mekong River Basin. The data were retrieved from the Lancang-Mekong Cooperation and Conflict Database (LMCCD) and double-checked with the literature review of the official documents and media sources related to Pöyry Report and Eyes on Earth Study.</span></p>

Beyond Perrault's experiments: repeatability, didactics and complexity

The naturalistic and philosophical studies conducted in the second half of the 17th century were crucial both for the birth of modern hydrological science and modern epistemology. Thanks to quantitative observations and to the new experiment-based scientific approach, the Sun was about to be fully recognized as the engine of the hydrological cycle. In this context of great vitality and rapid cultural changes, Pierre Perrault published his classical opus De l'origine des fontaines (On the origin of springs) in 1674. The opus presents a discussion on the origin of springs and contains the report of a set of experiments of water flow through a soil column, which may be considered the first of modern hydrology. In assessing the importance of Perrault's opus, we will discuss his epistemological relevance by looking at the novelty of his approach, at the repeatability of the experiments, at the intriguing didactic aspects for the modern teaching of hydrology and at his attitude in facing the complexity of hydrological processes. Perrault places himself in the context of a novel experimental epistemology. On the basis of our analyses he seems to be aware that the processes involved in the hydrological cycle and in soil hydrology are hardly reproducible by means of a controlled laboratory model. This circumstance put the modern scientific approach to a severe test at its very beginning. It is suggested that some of Perrault's epistemological and methodological reflections are precursors of the modern epistemology of complexity. Thus even if Perrault's conclusions followed an ancient opinion, his work is not only seminal for hydrology, but also helps to enlighten some features of the scientific revolution of the 17th century.

Deep Learning for Drought and Vegetation Health Modelling: Demonstrating the utility of an Entity-Aware LSTM

<p>Tools from the field of deep learning are being used more widely in hydrological science. The potential of these methods lies in the ability to generate interpretable and physically realistic forecasts directly from data, by utilising specific neural network architectures. </p><p>This approach offers two advantages which complement physically-based models. First, the interpretations can be checked against our physical understanding to ensure that where deep learning models produce accurate forecasts they do so for physically-defensible reasons. Second, in domains where our physical understanding is limited, data-driven methods offer an opportunity to direct attention towards physical explanations that are consistent with data. Both are important in demonstrating the utility of deep learning as a tool in hydrological science.</p><p>This work uses an Entity Aware LSTM (EALSTM; cf. Kratzert et al., 2019) to predict a satellite-derived vegetation health metric, the Vegetation Condition Index (VCI). We use a variety of data sources including reanalysis data (ERA-5), satellite products (NOAA Vegetation Condition Index) and blended products (CHIRPS precipitation). The fundamental approach is to determine how well we can forecast vegetation health from hydro-meteorological variables. </p><p>In order to demonstrate the value of this method we undertook a series of experiments using observed data from Kenya to evaluate model performance. Kenya has experienced a number of devastating droughts in recent decades. Since the 1970s there have been more than 10 drought events in Kenya, including droughts in 2010-2011 and 2016 (Haile et al 2019). The National Drought Monitoring Authority (NDMA) use satellite-derived vegetation health to determine the drought status of regions in Kenya.</p><p>First, we compared our results to other statistical methods and a persistence-based baseline. Using RMSE and R-squared we demonstrate that the EALSTM is able to predict vegetation health with an improved accuracy compared with other approaches. We have also assessed the ability of the EALSTM to predict poor vegetation health conditions. While better than the persistence baseline the performance on the tails of the distribution requires further attention.</p><p>Second, we test the ability of our model to generalise results. We do this by training only with subsets of the data. This tests our model’s ability to make accurate forecasts when the model has not seen examples of the conditions we are predicting. Finally, we explore how we can use the EALSTM to better understand the physical realism of relations between hydro-climatic variables embedded within the trained neural network. </p><p> </p><p>References:</p><p>Gebremeskel, G., Tang, Q., Sun, S., Huang, Z., Zhang, X., & Liu, X. (2019, June 1). Droughts in East Africa: Causes, impacts and resilience. Earth-Science Reviews. Elsevier B.V. https://doi.org/10.1016/j.earscirev.2019.04.015</p><p>Klisch, A., & Atzberger, C. (2016). Operational drought monitoring in Kenya using MODIS NDVI time series. Remote Sensing, 8(4). https://doi.org/10.3390/rs8040267</p><p>Kratzert, F., Klotz, D., Shalev, G., Klambauer, G., Hochreiter, S., & Nearing, G. (2019). Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets. Hydrology and Earth System Sciences, 23(12), 5089–5110. https://doi.org/10.5194/hess-23-5089-2019</p><p>Github Repository: https://github.com/esowc/ml_drought</p>