Hydrol. Earth Syst. Sci.

, 25, 4549–4565, 2021

https://doi.org/10.5194/hess-25-4549-2021
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Rainbow color map distorts and misleads research in hydrology –

guidance for better visualizations and science communication
Michael Stoelzle1 and Lina Stein2
1 Faculty of Environment and Natural Resources, University of Freiburg, Freiburg, Germany
2 Department of Civil Engineering, University of Bristol, Bristol, UK

Correspondence: Michael Stoelzle ([email protected])

Received: 27 February 2021 – Discussion started: 9 March 2021

Revised: 21 June 2021 – Accepted: 30 June 2021 – Published: 24 August 2021

Abstract. Nowadays color in scientific visualizations is stan- 1 Why does the rainbow color map distort and mislead
dard and extensively used to group, highlight or delineate scientific visualizations?
different parts of data in visualizations. The rainbow color
map (also known as jet color map) is famous for its appeal-
ing use of the full visual spectrum with impressive changes in Colorful visualizations are deeply integrated in science com-
chroma and luminance. Besides attracting attention, science munication. In hydrology, visualization of water fluxes like
has for decades criticized the rainbow color map for its non- precipitation, evapotranspiration, discharge or percolation
linear and erratic change of hue and luminance along the data and terms like green and blue water, humidity and aridity,
variation. The missed uniformity causes a misrepresentation or flood and drought are subjects of the daily hydrologists
of data values and flaws in science communication. The rain- work. Our presentation of patterns, relationships, composi-
bow color map is scientifically incorrect and hardly decod- tions, distributions and comparisons of multivariate datasets
able for a considerable number of people due to color vision is often multifaceted. And they are most often encoded with
deficiency (CVD) or other vision impairments. Here we aim color (Wong, 2011a). This is first of all reasonable as human
to raise awareness of how widely used the rainbow color map perception is dominated by visual perception (70 % com-
still is in hydrology. To this end, we perform a paper survey pared to 30 % by the other senses). The human eye can rec-
scanning for color issues in around 1000 scientific publica- ognize around 10 million unique colors but only 30 shades of
tions in three different journals including papers published grey (Kreit et al., 2013). Today computer software and freely
between 2005 and 2020. In this survey, depending on the available programming tools like R or Python simplify the
journal, 16 %–24 % of the publications have a rainbow color use of color and color gradients in color maps. The rise of
map and around the same ratio of papers (18 %–29 %) uses online-only journals reduced the necessity for a good per-
red–green elements often in a way that color is the only pos- ception of black–white printed graphs or papers. Although
sibility to decode the visualized groups of data. Given these colorful graphs and maps can be created with a few clicks,
shares, there is a 99.6 % chance to pick at least one visual the development of a compelling visualization is a complex
problematic publication in 10 randomly chosen papers from task.
our survey. To overcome the use of the rainbow color maps In terms of correct encoding, visual mappings such as po-
in science, we propose some tools and techniques focusing sition, length, angle, direction, area and volume rank higher
on improvement of typical visualization types in hydrologi- in efficiency and accuracy than color (e.g., Wong, 2010). In
cal science. We give guidance on how to avoid, improve and other words, the human eye is stronger in encoding data that
trust color in a proper and scientific way. Finally, we out- is mapped in a bar plot or scatterplot than in a colorful heat
line an approach how the rainbow color map flaws should be map. The encoding accuracy of color maps has primarily
communicated across different status groups in science. been criticized when it comes to the rainbow color map. This
color map uses all wavelengths of the visible spectrum be-
tween 380 and 750 nm, impresses with high lightness and

Published by Copernicus Publications on behalf of the European Geosciences Union.

4550 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

Figure 1. Comparison of different color maps along an arbitrary scale (0–6). The same delta changes in x values (denoted with +1) are not
represented uniformly in the rainbow or heated body color map due to unordered luminance. Alternatives are monochromatic, single-hue
color maps (greys, blues) or perceptually uniform designed multi-hue color maps like viridis or plasma. Diverging color maps highlight data
extremes and the data direction (e.g., positive/negative data values) if a meaningful midpoint is apparent in the data. Visualization inspired
by the literature (Crameri et al., 2020; Wong, 2011b).

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4551

chroma, encapsulates the most saturated colors, and hence rainbow color map as perceptually much slower and more
looks at first sight very appealing and eye-catching (Fig. 1). error-prone compared to single-hue color maps or perceptu-
In the past, the rainbow (or jet) color map or red and green ally uniform designed multi-hue color maps (Liu and Heer,
colors for sequential data were often the software standard, 2018).
having a wide use in the scientific communities and hence A thoughtful and scientifically correct color map should
also in publications. For example, in former versions of the allow for all types of dichromatic views (i.e., color vision de-
statistics software R (version 3.x) the pre-set color map used ficiency) and unambiguous perception of the displayed data.
black, red, green and dark blue as the first four colors and As it is now scientific standard and best practice to avoid
failed numerous colorblind checks. any rainbow or rainbow-derivate color map (Crameri et al.,
In general, there are two main reasons why the rainbow 2020), we want to challenge the use of rainbow color maps
color map in scientific visualization is “(still) considered in hydrological science by analyzing the status quo of rain-
harmful” (Borland and Taylor, 2007). First, color vision de- bow visualizations in hydrological and environmental publi-
ficiency (CVD) affects the perception of 8 %–10 % of the cations. In this study we evaluate the use of rainbow and red–
male and 0.4 %–0.5 % of the female population, depending green color use in the journal Hydrology and Earth System
on earth regions, and thus up to 4 % of the world’s popu- Sciences (HESS) in preprints and publications in 2020 and
lation (Geissbuehler and Lasser, 2013; Nuñez et al., 2018; their use over time (2005–2020). We then compare the re-
Pramanik et al., 2012). CVD shares are given for Caucasian sults to two other journals which cover different disciplines.
people and might be lower among other ethnic groups. The Finally, we discuss alternatives for using color overall in sci-
simultaneous and side-by-side use of red and green as in the entific publications and how to improve and trust the use of
rainbow color map obstructs an unbiased access to the vi- color.
sualization for these people. Putting an 8 % CVD ratio into
perspective, a Caucasian male team of one editor and two re-
viewers during a paper review has a chance up to 22.1 % that 2 Meta-analysis with paper survey
at least one person has a CVD (Wong, 2011b).
Secondly, the rainbow color map attracts attention but 2.1 How often is the rainbow color map used in
is weak in representing data in a scientifically correct way scientific visualizations?
(Fig. 1). This affects all people, even those with normal color
vision. The same Euclidean distances in mapping or the same There are discrepancies between theoretically known scien-
data ranges in continuous or binned variables are not equally tific standards and the de facto use of the rainbow color map.
represented by a rainbow color map (Crameri et al., 2020; A non-representative survey of presentations and posters at
Sharma and Trussell, 1997). Especially for data comparisons EGU 2018 (European Geosciences Union) found 60 % in-
over a wider distance in the color map, the distorted colors of cluded at least one rainbow scale figure (McNeall, 2018).
the rainbow impede reliable judgments (Liu and Heer, 2018). Compared to publications, visualizations are even more es-
Abrupt changes of lightness and saturation often lead to an sential for poster presentations and conference talks as less
unintended focus on some sections of the data range (Thyng time and text is available to present the research results. The
et al., 2016; Wong, 2011a). The high lightness of the yellow, appealing effect of rainbow color maps is often used as eye-
cyan or magenta segments in the rainbow color map makes it catcher along the poster walls. Due to the peer-review pro-
difficult to perceive a consistent color and data value ordering cess, we hypothesized that the ratio of rainbow color maps in
(Kovesi, 2015). Also, high and low values could be confused publications should be notably lower than 60 %. If there is a
if both are represented by reddish colors at the edges of the considerable number of scientific publications with rainbow
rainbow color map. The color map distorts the data repre- color maps, is there at least a decreasing tendency towards
sentation if the change in value is not visually commensurate fewer rainbow color maps in recent years?
with the change in color (Wong, 2010, 2011c). Discordant First, we examined all preprints that were published
false coloring may lead to visual errors up to 7.5 % of the to- in October 2020 (n = 36) in the journal Hydrology and
tal displayed data variation (Crameri et al., 2020). For exam- Earth System Sciences (HESS). We found 25 % of these
ple, research has also shown that replacing a rainbow color preprints having at least one graph or map with a rain-
map with a perceptually uniform color map could identify bow color map. Three of these rainbow preprints col-
hidden structures in mapping (Rogowitz et al., 1996). In per- ored 70 %–80 % of all figures with a rainbow color map.
ceptually uniform color maps, the delta change in color is Interestingly, the median author number of the rainbow
equal to delta change in data. Comparisons of rainbow color preprints was five, suggesting that rainbow-colored visu-
maps and perceptually uniform color maps in cartographic alizations are not necessarily seen as a critical issue dur-
mapping have demonstrated that rainbow colors can empha- ing manuscript preparation and internal submission pro-
size strong gradients where actually smooth data variation is cesses. We then consulted the author guidelines from the
apparent (Fig. 3 in Thyng et al., 2016). Empirical judgment journal (https://www.hydrology-and-earth-system-sciences.
of different quantitative color maps has hence identified the net/submission.html, last access: 29 July 2021) to check what

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4552 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

kind of color recommendations are given for the authors. We color or problems with readability of the graph. Only two
found in total two occurrences of the term “color” on the comments specifically address the issue a red–green color
web page. In the section “Figures and tables”, recommen- scale will have for some readers. In reaction, one published
dations for high-quality graphics are given with “For maps article changed the color scale to orange–blue instead of
and charts, please keep color blindness in mind and avoid the red–green. The other article changed the color map in most
parallel usage of green and red. For a list of color scales that graphs to an improved color map but also did change one
are illegible to a significant number of readers, please visit color map to rainbow color map. For a further three articles,
ColorBrewer 2.0”. the reviewer criticized readability of the graph. In reaction,
In a second step after the preprint analysis, we evaluated if one article changed a red–green coloring to red–black. The
the review process reduces the use of rainbow-colored visual- other two did not change their figures and continued the use
izations. We did that by screening in total 263 peer-reviewed of red–green or rainbow colors. This demonstrates that re-
papers published in HESS in the year 2020. To our knowl- viewers and editors are not sufficiently aware of this prob-
edge, no systematic review of rainbow color maps in envi- lem, which is also justified by comparable rainbow color map
ronmental journals exists so far. The journal guidelines of shares in preprints and peer-reviewed, published papers.
HESS also recommend to avoid green and red colors side by We then extended the survey for all papers published in
side in visualizations. We therefore classify the papers into HESS in 2015, 2010 and 2005 to better understand if there is
four groups also considering pure black–white papers: or was a tendency towards more or fewer rainbow-colored
visualizations in scientific publications (Fig. 2, Table A1;
A. black–white paper without use of any color,
https://github.com/modche/rainbow_hydrology, last access:
B. paper that has no rainbow-colored visualization or sup- 23 August 2021). Nearly half of the examined 800 papers
ports distinction with additional elements, (from 2005, 2010, 2015 and 2020) have been classified with
critical use of ambiguous or not colorblind-friendly color
C. paper that has at least one visualization with rainbow- maps (Fig. 2). Survey results indicated that the ratio of pa-
related coloring or use of green and red elements with- pers with rainbow or rainbow-related color maps has been
out a good chance to separate these elements, stable between 2010 and 2020 but markedly increased be-
tween 2005 and 2010. From 2005 to 2010 there was a clear
D. paper that has at least one rainbow-colored visualization
increase in color use (from 56 % to 82 %), and black and
(graph or map).
white papers dropped by 26 %. The survey of 797 papers
That means a graph with a red and green boxplot could be results in 9 % pure black and white papers; 47 % of chro-
classified as acceptable (class B) as often axis labels explain matic papers showed no color issues, and 44 % of all papers
the boxplot groups. A graph with two lines (red and green; have either used a rainbow color map in at least one visual-
see Fig. 6 “Original”) encoding continuous variables over ization and/or have embedded a visualization with red–green
time without any annotations more than the legend is clas- issues. Two cross checks with 30 and 50 randomly chosen
sified as rainbow-related (class C). If a paper has a rainbow papers led to minor deviations in color classification due to
color map visualization, then potential misuse of red/green is the personal judgment of our reviewer team (three people).
not further counted in our statistics. However, a high fraction of rainbow classified papers from
The majority of 201 (76 %) papers in 2020 have not the main survey were also classified as rainbow papers in the
included any rainbow-colored visualization. For 62 papers cross checks, with 86 % (6 out of 7 papers) and 92 % (12 out
(24 %), we found at least one graph or map that uses ex- of 13 papers) agreement.
plicitly the rainbow color map. In 60 papers (23 %) we clas-
sified at least one graph as “rainbow-related” (e.g., use of 2.2 Is the use of rainbow color maps a journal- or
the spectral color map) or identified red–green data encod- discipline-specific artifact?
ings without a good chance to distinguish different lines or
points. Summarizing these color issues, our survey shows To answer this question, we screened in total 200 additional
that around 47 % of the publications in HESS in 2020 had publications from different disciplines (e.g., environmental
visualizations that are not scientifically correct, not percep- science, biology) in the renowned journals Nature Scien-
tually uniform, and unaccessible or hard to access for around tific Reports and Nature Communications. For Nature Scien-
4 % of the readership due to color vision deficiency. This tific Reports, we looked at the top 100 of most downloaded
indicates that the awareness of misleading color choice is papers in the section “Earth Science” in 2019 (accessible
rather low during the publication process for both authors via https://www.nature.com/collections/agegihhehi, last ac-
and reviewers. This was further confirmed by evaluating re- cess: 29 July 2021). On the corresponding website, graph-
viewer comments of articles published in 2020. We searched ical thumbnails are given to preview the research findings.
for keywords “blind”, “color”, “colour”, “green” and “defi- Here we found 10 out of 100 thumbnails have rainbow color
ciency”. Of the 263 articles published, nine reviewer com- maps or rainbow-related coloring. Going into more detail we
ments (3.4 %) mention necessary improvements regarding scanned the 100 papers and found that 26 % of the papers

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4553

Figure 2. Proportion of surveyed papers without and with color issues. In total 797 peer-reviewed papers from Hydrology and Earth System
Sciences (HESS) with different shares for the years 2005, 2010, 2015 and 2020 were analyzed. Missing percent up to 100 is due to rounding.
In 2020 only papers that were published before 1 November are considered.

used rainbow color maps in at least one figure (class D) and 50 %–60 % of papers having at least one rainbow visualiza-
18 % have figures with potential red–green issues in color tion. Putting our survey results into perspective, when pick-
perception (class C). We also counted 6 % of the surveyed ing randomly 5 (10) papers one still has a 75.0 % (93.7 %)
papers in this journal as black and white papers (class A). chance to encounter at least one paper with a rainbow visu-
However, more than half of the investigated papers in Na- alization. If red–green issues are also considered, then the
ture Scientific Reports have been classified to be color-issue chance of at least one visual problematic paper in a selection
free (Fig. 3). In Nature Communications (survey of 100 pa- of 5 (10) papers is 94.4 % (99.6 %). Our survey suggests a
pers published between 4 and 6 November 2020), we found co-occurrence between the decline of black and white papers
no papers in class A (black–white), 55 % of papers without and the emergence of papers with color issues. For 2005 we
color issues (class B), 29 % of papers that used red and green find that 20 % of the papers published in HESS have color
(class C), and 16 % with a rainbow color map (class D) in at issues, and for 2020 we find that 47 % of papers have color
least one visualization (Fig. 3). issues. In the same time the share of black and white papers
Altogether, we surveyed 997 scientific papers from three dropped from 44 % to less than 1 % in 2020.
journals (published between 2005 and 2020, with 46 % of When analyzing the effect of color issue awareness among
papers from 2019 and 2020) and found 23.7 % of all pa- author teams, we clearly see that a higher number of au-
pers have at least one visualization colored with a rainbow thors did not necessarily lead to a lower share of color-
color map. The ratio of misused red–green color combina- issue-free papers (Fig. 4). Chromatic papers without color
tions is most likely even higher than reported here as red– issues had a 40 %–50 % share regardless of how many au-
green issues in rainbow papers are not separately counted in thors were included in these publications. Black and white
the statistics. However, our results revealed a considerably papers are mostly published as single-author papers or by
lower ratio of rainbow color maps compared to Borland and teams of two authors (27 % and 14 %, respectively). How-
Taylor (2007). Their survey from the 2001 through to 2005 ever, only 4 % of all papers are published with visualizations
IEEE Visualization Conference proceedings found around in black and white, and the share has been decreasing since

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4554 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

3.1 Avoid color – learning from black and white

visualizations

Taking inspiration from older papers with black and white

visualization is a valuable approach to identify potential im-
provement of colored visualizations. In former times technol-
ogy and/or computer software did not allow for the same use
of color in visualizations as today. Years ago, colored pages
in visualizations were also additionally charged by the pub-
lishers. Today color is often the first choice for data encod-
ing in visualizations (Wong, 2011a), but colors are also of-
ten used without any reason. Despite the fact that the human
eye can differentiate millions of colors, Stauffer et al. (2015)
stated that only a small number of different hues can be pro-
cessed for important classification tasks (search and distin-
guishing). Healey (1996) showed that only around seven dif-
ferent hues can be found accurately and rapidly on a map
or cartographic application. On maps, and also heat maps,
the neighboring colors and the distance between two colored
elements bias the perception of data variability (Brychtová
Figure 3. Share of papers with at least one visualization with red– and Çöltekin, 2017). If so, visualizations with extensive color
green issues or a rainbow color map across different journals. use should be revised to reduce the number of colors or re-
designed using other graphical encodings. Structure, hierar-
chy, clarity and completeness can instead be used to create
2005 (Fig. 2). The share of papers with rainbow color issue
an appealing look in a figure instead of color.
for single-author papers is around 9 %. In contrast, consid-
Some examples in the surveyed and other literature illus-
ering multiple-author publications, around 20 %–30 % of the
trate potential ways of doing this: visualization of model bi-
papers have rainbow color issues.
ases with a grey color gradient (Schaefli et al., 2005); black
We speculate that single-author publications are more of-
and white map shadings (Milly, 1994); response time dis-
ten composed by senior authors preferring simpler but often
tributions of different catchments with lines with various
clearer visualizations where color is not (primarily) needed to
greys, thicknesses, and line types (Roa-García and Weiler,
encode the data. Another feasible explanation is that in for-
2010); monthly regime curves of different climate models by
mer years papers were written by fewer authors than today.
lines with different line types and additional point symbols
Average author number per paper in our HESS paper survey
to highlight a specific baseline (Kingston and Taylor, 2010);
was 3.72 in 2005 and 5.06 in 2020. Text mining analysis of
monochromatic mappings and cumulative fluxes (Campbell
paper title terms (see Appendix, Fig. A1) suggests that often
et al., 2015); stacked bar charts with a sequential grey color
studies with spatial analyses or cartographic maps have an
scale (Sunyer et al., 2015); and various point shapes to rep-
above-average chance that a visualization with color issues
resent groups (Burden et al., 2019; Sunyer et al., 2015) or vi-
is embedded. Here papers with title terms such as “terres-
sualizations with direct labeling instead of legends (Hoellein
trial”, “map(s)” or “mapping”, “radar”, “satellite”, and “re-
et al., 2019).
gion” or “regional” have in between 73 % and 92 % of the
Hydrology as a science particularly uses line charts to il-
cases (Fig. A1) a rainbow-colored figure or a figure with red–
lustrate change over time, e.g., in streamflow analysis. We
green issues in the paper.
thus suggest to find visual inspiration from black and white
papers to demonstrate how color could be avoided or re-
3 Four steps to go beyond the rainbow color map duced. At the same time, specific aspects of a visualization
can be explicitly highlighted for the reader (Fig. 6). If a sin-
To overcome the need for the rainbow color map, we present gle technique is not sufficient to improve the visual state-
four suggestions to avoid, improve, trust and communicate ment, then a combination of techniques could be also fea-
color in scientific visualizations. The central questions for sible, e.g., lines with various widths and types, additional
these four steps are given in Fig. 5 as techniques to improve overlaid points on the lines, or direct labels to highlight spe-
the use of color and to communicate misuse of color. Kelle- cific lines such as the baseline or the mean (Fig. 6). Espe-
her and Braswell (2021) published a comprehensive intro- cially direct labeling could improve the clarity of the (line)
duction on how visualization challenges of environmental graphs leading to fewer cluttered graphs due to additional
data can be assessed in a broader sense. white space when the legend box is removed (Fig. 6e and f).
Text elements, rich in contrast, give guidance for the reader

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4555

Figure 4. Composition of color issues with a focus on number of authors. Labels of x axes show minimum and maximum values among the
categories. Each row adds up to 100 %.

with the information in the caption (Rougier et al., 2014).

Informative, story-telling graph titles may enhance data vi-
sualizations (Wanzer et al., 2021), although they are still un-
popular in the scientific community.
Besides improvements in single visualizations, splitting
figures and maps into different subplots (i.e., facets in
Fig. 6h) allows for multiple views on the story of a visualiza-
tion (Shoresh and Wong, 2012). Here the facets or sparklines
(Streit and Gehlenborg, 2015) replace different colors in a
single plot. A common technique is to present all data points
in all facets as a background data variation and then use color
Figure 5. Circular flow of central questions to improve and com- or point shapes to highlight specific data groups in the single
municate color in scientific visualizations. The different tasks are facets (Gnann et al., 2019, 2020). This multi-facetted view
discussed in the Sect. 3.1 to 3.4. Table 1 provides a checklist for the might be valuable for Budyko curve analyses, visualization
different tasks (properties, tools, etc.). of different model runs, catchment comparisons or to high-
light different distributions of data groups along one axis
(histograms, density plots, area charts), on two axes (scat-
terplots) or if grouping in stacked visualizations (bar or area
to comprehend the story and support people with low vision charts) is encoded by color. For example, dense scatterplots
to easily identify the major elements of the visualization. with a lot of overplotting, like storage–discharge plots in hy-
Scientists have the possibility to use other visual encod- drological recession analysis (Stoelzle et al., 2013) or in large
ings than color such as position (e.g., scatterplots), length sample hydrology, could profit from faceting as then data en-
(e.g., bar charts), or different point shapes or line widths coding is shifted from color to position.
(Fig. 6a–d) to increase the perception of data variation in Another important issue in hydrological science is heat
the visualization (Kelleher and Wagener, 2011; Wong, 2010). maps as they allow us to visualize a third variable in a two-
Furthermore, variation in transparency (Fig. 6c) could also dimensional coordinate system. Heat maps notably rely on
improve the clarity of visualization especially when data color encoding and thus need an appropriate color map and
points are plotted above each other like in large sample hy- a meaningful order of the data (categories) on rows and
drology or other scatterplots. If data encoding is shifted from columns. Rethinking the order of rows and columns in heat
color to other mappings such as position or line width, a tar- maps often reveals that an alphabetical or a chronological
geted use of one or two colors is a strong technique to high- order is not the best choice. Similarity and clustering of cat-
light specific parts of the visualization such as baselines, ex- egories may help out here (Gehlenborg and Wong, 2012a),
tremes, averages, or specific periods or regions (Fig. 6g). Im- and clustering can be done by splitting a single heat map into
proved visualization should be accompanied by elaborative multiple heat maps. As color perception in a heat map de-
figure captions (Fig. 6i), and all authors are asked to create pends on colors of neighboring cells, varying stroke color or
self-explanatory figures that are fully understandable solely

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4556 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

Figure 6. Recommendations to improve colorful line graphs or line graphs with red and green lines next to each other (steps 1 and 2).
Improvements can be achieved (steps 3 and 4) by adjustment of line width, type or brightness (a–c); adding additional points or labels to
lines (d–f); focusing on one specific line (g); using small multiples to allow for easier comparison (h); and ensuring that the figure is fully
self-explanatory with a precise and complete figure caption and an informative graph title (i). Additional, more advanced techniques (4.) are
colorblind-friendly color maps, increased contrast and dark(er) backgrounds (e.g., increasing contrast to highlight lines during a presentation)
to further improve (line) charts.

line width around the tiles will improve the data perception vision, high contrast and supportive text elements or point-
(Supplementary Fig. 3 in Crameri et al., 2020). Adding nu- ers are most important. Visualization should hence be im-
meric values as text to heat map tiles reduces the ambiguity proved with high text and element contrasts, annotations, the
of color perception. Parallel coordinate plots might be a valu- ease of horizontal labeling without line breaks, clear figure
able alternative to heat maps by shifting the data encoding structure with elaborated hierarchy and focus within the vi-
from color to position (Gehlenborg and Wong, 2012a). Mul- sualization (Tufte, 1983). Gestalt principles such as similar-
tivariate data can be split into several two-dimensional visu- ity, closure, proximity and common regions help to achieve
alizations to improve the clarity of the visualization (Gehlen- grouping and partitioning in graphs. The idea of data–ink ra-
borg and Wong, 2012c). tio (Tufte, 1983) could help to remove cluttered, non-data el-
But there is another very important aspect in advancing ements and allow for more white space as this helps to focus
data visualizations. Extensive use of color and perceptually on important parts of the visualization. Fewer, thinner or re-
ineffective color maps like the rainbow color map impede not moved grid lines also increase the data–ink ratio and sharpen
only people with CVD but also affect people with low or re- the view on the data.
duced vision. The World Health Organization (WHO) states In a second step, changing perspective from authors to
that “at least 2.2 billion people have a vision impairment” publishers, the community needs more advanced tools than
(World Health Organization, 2019). Although a lot of those static PDF or printed papers (Vandemeulebroecke et al.,
people have received professional eye care, a reduction of vi- 2019). If articles are more and more published in HTML for-
sual acuity could impede access to overloaded and cluttered mat, interactive visualizations could allow for multiple per-
visualizations with less contrast. It is important to recognize spectives by data zoom, selection and layering (e.g., Gehlen-
that the group of people with low or limited vision or visibil- borg and Wong, 2012c). Interactive visualizations give the
ity is much larger (up to 28 % of world’s population) than the possibility for people with low vision or CVD to select sup-
group of people with CVD. To visualize for people with low portive elements like tooltips during mouse hovering or by

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4557

highlighting selected elements and give direct labeling and In summary, finding an alternative for the rainbow color
annotations. Using alt attributes in HTML can be used to map can be seen as a relatively straightforward process
specify alternative text if a visualization cannot be rendered when two main aspects of color maps are considered. Firstly,
or perceived. Interactive elements on basis of HTML and there are many perceptually uniform color maps available
R packages are, for example, dygraphs (Vanderkam et al., that are well documented, professionally designed and do
2018), with sliders to select specific time periods in time se- consider people with CVD. As a starting point, the online
ries analysis or leaflet maps (Cheng et al., 2021), giving the tool ColorBrewer 2.0 (https://colorbrewer2.org, last access:
possibility to zoom into maps and to select different back- 29 July 2021) can be used to find appropriate, colorblind-
ground layers (i.e., leading to higher contrast for people with friendly color maps. Such color maps are also often avail-
vision impairment). Authors can also accompany their data able in the visualization software or programming languages
analysis in a paper with an online available data dashboard for example the viridis (Garnier, 2018) or scico (Pedersen
such as Shiny apps (Chang et al., 2020) for further data ex- and Crameri, 2020) packages in R or the seaborn library in
ploration. Such efforts could be also beneficial for people Python (Waskom, 2021). The MATLAB software uses the
with other limited (cognitive) capabilities (i.e., low visual- “parula” color map as a default, but also other color maps
ization literacy or blind people) as there are possibilities of can be set up. Regardless of the used software, a thought-
speech- or touch-based interaction with non-static visualiza- ful color map should be perceptually uniform, colorblind
tions (Lee et al., 2020). friendly, strong in greyscale conversion and – if possible –
pretty and appealing to attract the reader (Fig. 7). Famous ex-
amples are the Okabe color map (Okabe and Ito, 2008), the
3.2 Improve color – what are alternatives to the
viridis color map (Garnier, 2018) or the recently published
rainbow color map?
scientific color maps by Crameri (2020).
Secondly, a simple guidance on the necessity for color
Although black and white visualizations could inspire a should be considered. The main uses of color are within these
thoughtful revision of colored figures or maps, removing all four categories:
color is not always the best choice (Kelleher and Braswell,
A. distinguish categories (i.e., each color is a category),
2021). Depending on data dimensions, the use of color is
sometimes unavoidable. Typically, we want to use color to B. visualize sequential data values (i.e., each color is a nu-
convey data and also to create a figure or map that looks ap- meric value),
pealing. Crameri et al. (2020) presented a thoughtful decision
tree explaining how color depending on data types should be C. visualize diverging data values (i.e., each color is a nu-
used in visualizations. They differentiated between the di- meric value and data have a direction and a meaningful
rection of color gradients to encode higher or lower values midpoint or centric value), and
according to the chosen background color (light or dark).
D. highlight some categories as special case of category
Dark(er) backgrounds have recently gained more and more
A; i.e., a few categories have a color and the remaining
attention in the visualization community as a possibility to
ones have a grey gradient.
increase the contrast of visualizations (Crameri et al., 2020).
Although dark figure backgrounds make for an unusual sight Moreover, there are also special cases of categories A–D
in articles, the increase in contrast might be appropriate for such as multi-sequential color maps (e.g., bathymetry and to-
presentations, helping people with vision impairments. If a pography with a centric value but not diverging) and circu-
single-hue color map such as the blue color map in Fig. 1 lar/cyclic color maps (e.g., river orientation, with repeating
is not sufficient to encode data by color, then blended-hue colors, e.g., 0 and 360◦ ) as outlined in Crameri et al. (2020).
color maps are a feasible solution. For example, to visualize We recommend as a first step to assign used data to one
depth below and elevation above sea level a combination of of the four categories A to D. If this is not possible, color
two monochromatic color maps with a reasonable midpoint might not be the best way to convey the data. Our survey re-
or breakpoint can be used (Gehlenborg and Wong, 2012b). vealed that especially the cases B and C were often mixed
If small data variations in a continuous color map are not up or used color in the wrong way (Fig. 7). Examples of
needed, then a discrete or binned color map with fewer but how to find appropriate color maps based on data types can
more distinguishable colors support faster decoding of color be found in the literature (Coalter, 2020; Light and Bartlein,
and data variation. Thoughtful breaks for color binning and 2004; Zeileis et al., 2020; Crameri et al., 2020). Empirical
better boundaries among important data ranges or regions assessment of quantitative color maps has also shown that di-
give guidance for the reader. Here the proximity and orien- verging color maps (e.g., blue–white–orange) are slower and
tation of the legend could also help to gain undistorted allo- more error-prone during data encoding compared to single-
cation of the presented data range. Depending on the degree hue color maps (Liu and Heer, 2018). Here comparisons of
of data variability on x and y axes, the legend could be hori- data values across the white midpoint are critical. For exam-
zontally or vertically aligned to the graph. ple, there is no need to have diverging color maps to visualize

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4558 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

Figure 7. Comparison of mapping with rainbow (first column) and perceptually uniform color maps (third column) for an arbitrary variable
across the counties of the US state of Texas. For each map, the vision for people with deuteranopia (red–green blindness) is given (second
and fourth columns). With the rainbow color map, values around 60 and 20 stick out (a, c) and greenish or yellowish colors dominate a
wide range of data variation. Compared to that, uniform and colorblind-friendly color maps (b, d) support a more exact perception of data
variation and the extreme values. White strokes between the map elements (b, d) increase the data–ink ratio. Maps with emulation of color
vision deficiency were created with the R package colorblindr.

correlation coefficient between 0 and 1, but they may be use- color choice in scientific visualizations. Firstly, by learning
ful for ranges between −1 and 1. However, diverging color from surveys testing the trustworthiness of colors and differ-
maps are meaningful and efficient as long they have a mean- ent color maps. Examples are the Marie Skáodowska-Curie
ingful midpoint or centric value (often zero, mean value or (Crameri et al., 2020) or the Which Blair project (Rogowitz
50 % value) and are used to highlight the direction of data and Kalvin, 2001), where different color palettes are mapped
or the data extremes, where the midpoint representation is to well-known photos to highlight the effect of different color
damped, for example, with light grey color. choice. These studies showed with experiments that color
Another important point is to know typical color maps of maps not based on a monotonically increasing luminance
specific types of visualization or in specific research areas component produced no positive rating scores (Liu and Heer,
(Gomis and Pidcock, 2018). For hydrology or environmen- 2018). Considering these findings, tools like ColorBrewer
tal science, the use of specific diverging color scales such as 2.0 (https://colorbrewer2.org, last access: 29 July 2021) give
red/blue for temperature or green/brown for aridity can also color advice on what kind of single- or multi-hue color maps
function as a first and easy decodable signal for the reader to can be used for sequential, diverging or qualitative data and
give advice on what is actually shown in the presented fig- let the user explicitly filter for colorblind safe and printer
ure or map (Fig. 7). Text elements or pointers will support friendly color maps (Coalter, 2020; Harrower and Brewer,
colored data decoding for people with low vision. Such the- 2003). With the R package colorspace (Zeileis et al., 2020),
matic and discipline-specific color choices aligned with peer a set of over 80 color palettes can be visualized and com-
discussions should be part of a thoughtful visualization de- pared among each other. In the R package scico (Pedersen
velopment within each author team. and Crameri, 2020) users will easily find palettes that em-
bed common colors for visualization in their disciplines (see
3.3 Trust color – tools to check for colorblind-friendly Sect. 3.2). The R package RColorBrewer (Neuwirth, 2012)
visualizations provides color maps for visualizations and offers the possi-
bility to use only colorblind-friendly color maps.
“In perceptual-uniformity we trust!” (Crameri, 2017). There Secondly, when a visualization is ready for publication au-
are mainly two possibilities to increase the trustworthiness of thors should aim for testing the specific figure or map to

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4559

Table 1. Checklist to improve color encodings in data visualizations. Tasks (1–4) are assigned to different roles during the publication
process: author, co-author, reviewer, editor, journal and audience. Single actions are classified in recommended (R) and/or advanced (A). All
URLs in this table were last accessed on 29 July 2021.

Role Level Action Reference and further reading

1. Avoid color
Author R Revise graph type to remove or reduce color encodings; Wong (2011a)
R Add second data encoding (e.g., point shapes) to resolve color Kelleher and Wagener (2011)
ambiguity;
A Use subgraphs or facets to focus attention and to reduce dimen-
sionality;
A Improve data–ink ratio by removing clutter (e.g., grid lines, Tufte (1983); Gomis and Pidcock (2018)
background colors).
2. Improve color
Author R Use the right color map according to data type. Be sure that Crameri et al. (2020)
sequential and diverging color map is not confused!
R Replace rainbow color map using https://colorbrewer2.org or Wong (2011c, b)
software; R and Python tools: Sect. 3.2
A If customized color map is needed, consider professional tools http://vrl.cs.brown.edu/color
and CVD accessibility; https://coolors.co
R/A Use discipline- and variable-specific color encoding. Clarify Gomis and Pidcock (2018)
and simplify graph with annotations, direct labeling, binned
color map and use color to highlight elements.
3. Trust color
Author R/A Inform yourself about trustworthy color palettes, e.g., Marie Crameri et al. (2020); Rogowitz and Kalvin
Co-author Skáodowska-Curie comparison and Which Blair project; (2001);
R Check your own work: run a CVD emulation to ensure accessi- see Sect. 3.2 for a list of tools
bility of color; https://contrastchecker.com
Check color contrast compliance or greyscale emulation.
4. Communicate color
Author A Add color statement: visualizations have been tested to be ac-
cessible for people with CVD (in papers, in presentations);
A Consider adding a greyscale version of a visualization to the
supplement;
Reviewer R Advise others on accessibility and potential deficiency of color Crameri et al. (2020)
Editor maps;
Consider color quality as review and decision criteria;
Journal R Provide clear color guidance including best practices on journal
home page (i.e., in obligations for referees and editors, in the
review criteria list);
Audience R Raise awareness of color issues with your colleagues and other Marie Skáodowska-Curie poster,
scientists; Crameri et al. (2020)
Give (friendly) criticism of rainbow color maps and propose
alternatives.

see how trustworthy the used color map actually is. With ternet without using a specific software, e.g., https://www.
the R package colorblindr (McWhite and Wilke, 2021), var- color-blindness.com (last access: 29 July 2021) or http://
ious types of color blindness are simulated for production- hclwizard.org (last access: 29 July 2021). Figure files can
ready R visualizations (Fig. 7). Open-source software ap- be uploaded to compare different visions such as normal,
plications for smartphones and computers offer a livestream deuteranopia, tritanopia or monochromacy. To improve visu-
of color-blindness emulation via camera or screen capture alizations for people with other vision impairments, online
(e.g., https://github.com/michelf/sim-daltonism/, last access: tools like https://contrastchecker.com (last access: 29 July
29 July 2021). Such emulations are also available on the in- 2021) help to test color contrast compliance or to perform

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4560 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

greyscale emulation. As a last step the R package paletteer figures and mappings in a fair and constructive way, propos-
(Hvitfeldt, 2021) offers the possibility to produce some de- ing methods to improve or avoid the (rainbow) coloring. Al-
scriptive statistics on a color map or publication-ready figure though the rainbow color map has more or less a tradition
and to run routines that optimize the color map, e.g., avoid in various hydrological subdisciplines (e.g., in visualizations
colors that are appearing too similar. Finally, colorblind- of water velocities, heat or solute transport as well as carto-
friendly color maps are not the answer to everything. Au- graphic maps in general), we especially encourage the net-
thors should still check for sufficient design of visualization works of young scientists to take responsibility for visualiza-
elements as point sizes are often too small, line widths are tions with valid color maps and a clear undistorted message.
often too thin or overplotting impedes a full view of the data. Communication of rainbow flaws should take place in all ar-
eas of science: during lectures, with colleagues, in network
3.4 Communicate (rainbow) color – what should meetings, as feedback for presentations, as a conference at-
scientists and publishers do? tendee or paper reviewer but also as a journal editor, senior
scientist or professor. The anti-rainbow Marie Skáodowska-
Literature review and our paper survey of color maps in sci- Curie poster from Crameri et al. (2020) is freely available and
entific publications suggest that there is a considerable dis- could be a communication starter at the wall near the coffee
crepancy between what science knows about the rainbow machine of your institute.
color map and what scientists do about it. The presented pa-
per survey shows this discrepancy. Around 46 % of published
HESS papers in 2020 have color issues (Fig. 2), but at the 4 Conclusions
same time only less than 4 % of all reviewer comments see
the choice of color as an issue before publication. From a The rainbow color map attracts attention but distorts and mis-
knowledge perspective, the rainbow color map distorts a cor- leads scientific visualizations. Major rainbow pitfalls are the
rect representation of the data variation. Thus, a reliable sci- non-linear data encoding, steps and disorder in luminance,
entific communication is not possible with the rainbow color and minor perceptual accessibility for people with CVD or
map. This knowledge is not reflected in the submission pro- other vision impairments (Figs. 1 and 7). Here we inves-
cess of many journals, although the journals have been re- tigated the use of rainbow color maps in around 1000 pa-
quested to raise author’s awareness about CVD accessibility pers in different environmental journals and found that the
issues (Albrecht, 2010). If the extensive use of rainbow color misleading rainbow color map or red–green color issues are
maps in science continues, then also journalists and (social) present in around 44 % of all papers (Fig. 2). We found no
media will most likely continue to circulate those rainbow journal-specific differences in the use of the rainbow color
figures and maps (Moreland, 2016). Today rainbow thumb- maps (Fig. 3). Compared to the knowledge about the flaws of
nails appear in graphical abstracts, as thumbnails on journal the rainbow color map, this share is alarmingly high. More-
websites or as screenshots in paper announcements on Twit- over, our hypothesis that rainbow color maps are on the de-
ter. This suggests for a broader audience that the rainbow cline could not be confirmed. Color issues in papers remained
color map is state of the art and scientifically correct as the constant or even increased between 2005 and 2020 (Fig. 2).
visualizations have been produced by scientists with a high Multi-author papers are not less prone to (rainbow) color is-
reputation in the public opinion. For young(er) scientists, the sues, even though more people could weigh in against inac-
high reputation of scientific journals justifies rainbow color cessible visualizations (Fig. 4). Analysis of reviewer com-
maps as appropriate for their own scientific work. ments highlights that the awareness of those issues is alarm-
To leave this vicious circle, a major effort in science com- ingly low during the review process.
munication is needed. Authors can actively state in their pa- Our survey indicates that past campaigns to banish the
pers that no rainbow color maps are used and visualizations use of rainbow color maps were not sufficient. We strongly
have been subjected to a color accessibility check. Publish- recommend that this issue should be raised across the hy-
ers and editorial teams should review graphical abstracts and drologic community. It will take students, researchers, lec-
summary thumbnails for rainbow color maps as authors tend tures, professors, editors, reviewers and publishers to ban-
to use rainbow-colored figures from the paper to attract read- ish the rainbow color map, as well as simultaneous red and
ers’ attention on websites with paper previews. Journal’s au- green usage, to make visualizations accessible for all read-
thor guidelines should specifically advise against the use of ers and to insist on correct data representation. As a guide
rainbow and red–green scales. Editors and reviewers should we presented manifold visual techniques on how to avoid,
ask for revisions of rainbow figures and should be more re- improve, trust and communicate color in data visualizations
lentless here. At scientific conferences, short courses for im- (Sect. 3.1–3.4, Table 1). Such guidance is given with a fo-
proved data analysis or environmental visualizations should cus on important graph types in hydrology to attenuate the
raise awareness for the rainbow color map topic, especially role and risks of color use in data encoding. Visualizations
for young(er) scientists. All of this is not about blaming the could not only be improved for people with CVD but should
authors of rainbow visualizations but to clearly criticize those be drafted with more care in terms of less exclusive data en-

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4561

coding by color (Fig. 6). Such efforts could also bring ad-
vantages for a much larger group of people with low vision
or vision impairments if more focus is given to visualizations
with less clutter, higher contrast and supportive graphical el-
ements like annotations.

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4562 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

Appendix A

Table A1. Survey of papers in the Hydrology and Earth System Sciences journal. Classification is done based on expert judgment. Missing
percent up to 100 is due to rounding. In 2020 only papers that were published before 1 November are considered.

Year 2005 2010 2015 2020 Total

Class Description Number of papers 54 191 289 263 797
A Pure black–white 44 % 18 % 4% <1% 9%
B Chromatic without color issues 35 % 41 % 48 % 53 % 47 %
C Red–green color issues 9% 16 % 22 % 23 % 20 %
D Rainbow color map issues 11 % 25 % 26 % 24 % 24 %

Figure A1. Text mining analysis finding 27 groups of papers sharing the same title term and having an above-average share of color issues.
A paper title is only considered for analysis if the group of papers with the specific term in the title has more than 44 % of papers with color
issues (above the average of 797 papers). Asterisk (*) acts as a wildcard in the regular expressions during text mining, e.g., with map* also
the terms maps and mapping are searched. Listed terms have a minimum of five letters (exception map* and snow*) and occur in at least
10 individual papers (exception bias correct* with 7 papers). A specific paper could be part of multiple paper groups. More generic words
like effect, approach, change, based, water or model and country names like China or France were excluded by expert judgment. Analysis
is based on all 797 surveyed papers from the journal Hydrology and Earth System Sciences (2005–2020). The median author number across
all groups of papers with the same title term ranges between 3.0 (map* and seasonal*) and 5.0 (evapotranspiration, Mediterranean, region*,
and soil moist*).

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4563

Data availability. Data of the paper survey and sample code snip- havior, Sci. Rep.-UK, 9, 4253, https://doi.org/10.1038/s41598-
pets are available online at https://doi.org/10.5281/zenodo.5145746 019-40396-x, 2019.
(Stoelzle, 2021a) and as Kaggle Notebooks for exploration at Campbell, J. M., Jordan, P., and Arnscheidt, J.: Using high-
https://www.kaggle.com/modche/rainbow-papersurvey-hydrology resolution phosphorus data to investigate mitigation measures in
(Stoelzle, 2021b). headwater river catchments, Hydrol. Earth Syst. Sci., 19, 453–
464, https://doi.org/10.5194/hess-19-453-2015, 2015.
Chang, W., Cheng, J., Allaire, J., Xie, Y., and McPherson, J.:
Author contributions. MS developed the research idea and de- shiny: Web Application Framework for R, available at: https:
signed the paper survey. MS and LS carried out the paper survey //cran.r-project.org/web/packages/shiny/index.html (last access:
and analyzed the results. MS wrote the article and designed the fig- 30 July 2021), 2020.
ures with contributions from LS. Cheng, J., Karambelkar, B., and Xie, Y.: leaflet: Create Interactive
Web Maps with the JavaScript “Leaflet” Library, available at:
https://cran.r-project.org/web/packages/leaflet/index.html, last
Competing interests. The authors declare that they have no conflict access: 30 July 2021.
of interest. Coalter, J.: ColorBrewer 2.0 and the Rainbow: Using Color Tools to
Choose Appropriate Color Schema for your Data Visualization,
ISTL, 94, 1–12, https://doi.org/10.29173/istl63, 2020.
Crameri, F.: The Rainbow Colour Map (repeatedly) considered
Disclaimer. Publisher’s note: Copernicus Publications remains
harmful, EGU Blogs, available at: https://blogs.egu.eu/divisions/
neutral with regard to jurisdictional claims in published maps and
gd/2017/08/23/the-rainbow-colour-map/, (last access 30 July
institutional affiliations.
2021), 2017.
Crameri, F.: Scientific Colour Maps, Zenodo,
https://doi.org/10.5281/zenodo.1243862, 2020.
Acknowledgements. Robin Schwemmle helped to evaluate the Crameri, F., Shephard, G. E., and Heron, P. J.: The misuse of
color issue classification, and Angelika Kuebert suggested R pack- colour in science communication, Nat. Commun., 11, 5444,
ages for colorblind-friendly color mapping. Thanks are due to https://doi.org/10.1038/s41467-020-19160-7, 2020.
Robin Schwemmle and Markus Weiler, whose comments greatly Garnier, S.: viridis: Default Color Maps from “matplotlib”, avail-
helped to improve the visualizations and the paper. Some visual- able at: https://cran.r-project.org/web/packages/viridis/index.
izations are inspired by the book Fundamentals of Data Visual- html (last access: 30 July 2021), 2018.
ization by Claus O. Wilke (online available on https://clauswilke. Gehlenborg, N. and Wong, B.: Heat maps, Nat. Methods, 9, 213–
com/dataviz/, last access: 30 July 2021). Authors also thank Fabio 213, https://doi.org/10.1038/nmeth.1902, 2012a.
Crameri, Thorsten Wagener and an anonymous reviewer for their Gehlenborg, N. and Wong, B.: Mapping quantitative data to color,
comments, which have helped to improve the paper. A community Nat. Methods, 9, 769–769, https://doi.org/10.1038/nmeth.2134,
comment from J. C. Refsgaard is also appreciated. 2012b.
Gehlenborg, N. and Wong, B.: Power of the plane, Nat. Methods, 9,
935–935, https://doi.org/10.1038/nmeth.2186, 2012c.
Financial support. This open-access publication was funded Geissbuehler, M. and Lasser, T.: How to display data
by the University of Freiburg. by color schemes compatible with red-green color
perception deficiencies, Opt. Express, 21, 9862,
https://doi.org/10.1364/OE.21.009862, 2013.
Review statement. This paper was edited by Jim Freer and reviewed Gnann, S. J., Woods, R. A., and Howden, N. J. K.: Is There a
by Fabio Crameri, Thorsten Wagener and one anonymous referee. Baseflow Budyko Curve?, Water Resour. Res., 55, 2838–2855,
https://doi.org/10.1029/2018WR024464, 2019.
Gnann, S. J., McMillan, H., Woods, R. A., and Howden, N. J.
K.: Including Regional Knowledge Improves Baseflow Signa-
ture Predictions in Large Sample Hydrology, Water Resour. Res.,
References 57, e2020WR028354, https://doi.org/10.1029/2020WR028354,
2020.
Albrecht, M.: Color blindness, Nat. Methods, 7, 775–775, Gomis, M. I. and Pidcock, R.: IPCC Visual Style Guide for Au-
https://doi.org/10.1038/nmeth1010-775a, 2010. thors, availabe at: https://www.ipcc.ch/site/assets/uploads/2019/
Borland, D. and Taylor, R.: Rainbow Color Map (Still) 04/IPCC-visual-style-guide.pdf (last access: 30 July 2021), 28,
Considered Harmful, IEEE Comput. Graph., 27, 14–17, 2018.
https://doi.org/10.1109/MCG.2007.323435, 2007. Harrower, M. and Brewer, C. A.: ColorBrewer.org: An Online Tool
Brychtová, A. and Çöltekin, A.: The effect of spatial distance on for Selecting Colour Schemes for Maps, Cartogr. J., 40, 27–37,
the discriminability of colors in maps, Cartogr. Geogr. Inf. Sc., https://doi.org/10.1179/000870403235002042, 2003.
44, 229–245, https://doi.org/10.1080/15230406.2016.1140074, Healey, C. G.: Choosing effective colours for data visualiza-
2017. tion, in: Proceedings of Seventh Annual IEEE Visualiza-
Burden, C. M., Morgan, M. O., Hladun, K. R., Amdam, G. V., tion ’96, Seventh Annual IEEE Visualization ’96, San Fran-
Trumble, J. J., and Smith, B. H.: Acute sublethal exposure to
toxic heavy metals alters honey bee (Apis mellifera) feeding be-

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021

4564 M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology

cisco, CA, USA, 27 October–1 November 1996, 263–270, Nuñez, J. R., Anderton, C. R., and Renslow, R. S.: Optimizing
https://doi.org/10.1109/VISUAL.1996.568118, 1996. colormaps with consideration for color vision deficiency to en-
Hoellein, T. J., Shogren, A. J., Tank, J. L., Risteca, P., and Kelly, able accurate interpretation of scientific data, PLoS ONE, 13,
J. J.: Microplastic deposition velocity in streams follows patterns e0199239, https://doi.org/10.1371/journal.pone.0199239, 2018.
for naturally occurring allochthonous particles, Sci. Rep.-UK, 9, Okabe, M. and Ito, K.: Color universal design (CUD) – how to make
3740, https://doi.org/10.1038/s41598-019-40126-3, 2019. figures and presentations that are friendly to colorblind people,
Hvitfeldt, E.: paletteer: Comprehensive Collection of Color available at: https://jfly.uni-koeln.de/color/ (last access: 30 July
Palettes, available at: https://cran.r-project.org/web/packages/ 2021), 2008.
paletteer/index.html, last access: 30 July 2021. Pedersen, T. L. and Crameri, F.: scico: Colour Palettes. Based
Kelleher, C. and Braswell, A.: Introductory overview: Recommen- on the Scientific Colour-Maps, available at: https://github.com/
dations for approaching scientific visualization with large envi- thomasp85/scico (last access: 30 July 2021), 2020.
ronmental datasets, Environmental Modelling & Software, 143, Pramanik, T., Khatiwada, B., and Pandit, R.: Color vision deficiency
105113, https://doi.org/10.1016/j.envsoft.2021.105113, 2021. among a group of students of health sciences, Nepal Med. Coll.
Kelleher, C. and Wagener, T.: Ten guidelines for effective data visu- J., 14, 334–336, 2012.
alization in scientific publications, Environ. Modell. Softw., 26, Roa-García, M. C. and Weiler, M.: Integrated response and transit
822–827, https://doi.org/10.1016/j.envsoft.2010.12.006, 2011. time distributions of watersheds by combining hydrograph sep-
Kingston, D. G. and Taylor, R. G.: Sources of uncertainty in cli- aration and long-term transit time modeling, Hydrol. Earth Syst.
mate change impacts on river discharge and groundwater in a Sci., 14, 1537–1549, https://doi.org/10.5194/hess-14-1537-2010,
headwater catchment of the Upper Nile Basin, Uganda, Hydrol. 2010.
Earth Syst. Sci., 14, 1297–1308, https://doi.org/10.5194/hess-14- Rogowitz, B. E. and Kalvin, A. D.: The “Which Blair project”: a
1297-2010, 2010. quick visual method for evaluating perceptual color maps, in:
Kovesi, P.: Bad Colour Maps Hide Big Features and Cre- Proceedings Visualization, 2001, VIS ’01, IEEE Visualization
ate False Anomalies, ASEG Extended Abstracts, 2015, 1–4, 2001, October 21 - October 26, 2001, San Diego, CA, USA, 183–
https://doi.org/10.1071/ASEG2015ab107, 2015. 556, https://doi.org/10.1109/VISUAL.2001.964510, 2001.
Kreit, E., Mäthger, L. M., Hanlon, R. T., Dennis, P. B., Naik, R. Rogowitz, B. E., Treinish, L. A., and Bryson, S.: How
R., Forsythe, E., and Heikenfeld, J.: Biological versus electronic Not to Lie with Visualization, Comput. Phys., 10, 268,
adaptive coloration: how can one inform the other?, J. R. Soc. https://doi.org/10.1063/1.4822401, 1996.
Interface, 10, 20120601, https://doi.org/10.1098/rsif.2012.0601, Rougier, N. P., Droettboom, M., and Bourne, P. E.: Ten Simple
2013. Rules for Better Figures, PLoS Comput. Biol., 10, e1003833,
Lee, B., Choe, E. K., Isenberg, P., Marriott, K., Stasko, J., https://doi.org/10.1371/journal.pcbi.1003833, 2014.
and Rhyne, T.-M.: Reaching Broader Audiences With Schaefli, B., Hingray, B., Niggli, M., and Musy, A.: A conceptual
Data Visualization, IEEE Comput. Graph., 40, 82–90, glacio-hydrological model for high mountainous catchments,
https://doi.org/10.1109/MCG.2020.2968244, 2020. Hydrol. Earth Syst. Sci., 9, 95–109, https://doi.org/10.5194/hess-
Light, A. and Bartlein, P. J.: The end of the rainbow? Color schemes 9-95-2005, 2005.
for improved data graphics, EOS T. Am. Geophys. Un., 85, 385, Sharma, G. and Trussell, H. J.: Digital color imaging, IEEE T.
https://doi.org/10.1029/2004EO400002, 2004. Image Process., 6, 901–932, https://doi.org/10.1109/83.597268,
Liu, Y. and Heer, J.: Somewhere Over the Rainbow: An 1997.
Empirical Assessment of Quantitative Colormaps, in: Pro- Shoresh, N. and Wong, B.: Data exploration, Nat. Methods, 9, 5–5,
ceedings of the 2018 CHI Conference on Human Factors https://doi.org/10.1038/nmeth.1829, 2012.
in Computing Systems – CHI ’18, the 2018 CHI Con- Stauffer, R., Mayr, G. J., Dabernig, M., and Zeileis, A.: Somewhere
ference, Montreal QC, Canada, 21–26 April 2018, 1–12, over the rainbow: How to make effective use of colors in mete-
https://doi.org/10.1145/3173574.3174172, 2018. orological visualizations, B. Am. Meteorol. Soc., 96, 203–216,
McNeall, D.: How many rainbows at EGU 2018?, Bet- 2015.
ter Figures, available at: https://betterfigures.org/2018/04/16/ Stoelzle, M.: Rainbow color map distorts and misleads re-
how-many-rainbows-at-egu-2018/ (last access: 30 July 2021), search in hydrology – guidance for better visualizations
2018. and science communication (v1.0), Zenodo [data set],
McWhite, C. D. and Wilke, C. O.: colorblindr: Simulate colorblind- https://doi.org/10.5281/zenodo.5145746, 2021a.
ness in R figures, available at: https://github.com/clauswilke/ Stoelzle, M.: Rainbow color map distorts and misleads research in
colorblindr, last access: 30 July 2021. hydrology – guidance for better visualizations and science com-
Milly, P. C. D.: Climate, soil water storage, and the average munication, Kaggle Notebook, available at: https://www.kaggle.
annual water balance, Water Resour. Res., 30, 2143–2156, com/modche/rainbow-papersurvey-hydrology, last access: 29
https://doi.org/10.1029/94WR00586, 1994. July 2021b.
Moreland, K.: Why We Use Bad Color Maps and What Stoelzle, M., Stahl, K., and Weiler, M.: Are streamflow recession
You Can Do About It, Electronic Imaging, 2016, 1– characteristics really characteristic?, Hydrol. Earth Syst. Sci., 17,
6, https://doi.org/10.2352/ISSN.2470-1173.2016.16.HVEI-133, 817–828, https://doi.org/10.5194/hess-17-817-2013, 2013.
2016. Streit, M. and Gehlenborg, N.: Temporal data, Nat. Methods, 12,
Neuwirth, E.: RColorBrewer: ColorBrewer Palettes, available 97–97, https://doi.org/10.1038/nmeth.3262, 2015.
at: https://cran.r-project.org/web/packages/RColorBrewer/index. Sunyer, M. A., Hundecha, Y., Lawrence, D., Madsen, H., Willems,
html (last access: 30 July 2021), 2012. P., Martinkova, M., Vormoor, K., Bürger, G., Hanel, M., Kri-

Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021 https://doi.org/10.5194/hess-25-4549-2021

M. Stoelzle and L Stein: Rainbow color map distorts and misleads research in hydrology 4565

aučiūnienė, J., Loukas, A., Osuch, M., and Yücel, I.: Inter- Wanzer, D. L., Azzam, T., Jones, N. D., and Skousen, D.: The role
comparison of statistical downscaling methods for projection of of titles in enhancing data visualization, Eval. Program Plann.,
extreme precipitation in Europe, Hydrol. Earth Syst. Sci., 19, 84, 101896, https://doi.org/10.1016/j.evalprogplan.2020.101896,
1827–1847, https://doi.org/10.5194/hess-19-1827-2015, 2015. 2021.
Thyng, K., Greene, C., Hetland, R., Zimmerle, H., and DiMarco, Waskom, M.: seaborn: statistical data visualization, JOSS, 6, 3021,
S.: True Colors of Oceanography: Guidelines for Effective https://doi.org/10.21105/joss.03021, 2021.
and Accurate Colormap Selection, Oceanography, 29, 9–13, Wong, B.: Design of data figures, Nat. Methods, 7, 665–665,
https://doi.org/10.5670/oceanog.2016.66, 2016. https://doi.org/10.1038/nmeth0910-665, 2010.
Tufte, E. R.: The visual display of quantitative information, CT Wong, B.: Avoiding color, Nat. Methods, 8, 525–525,
Graphics, Cheshire, 1983. https://doi.org/10.1038/nmeth.1642, 2011a.
Vandemeulebroecke, M., Baillie, M., Margolskee, A., and Magnus- Wong, B.: Points of view: Color blindness, Nat. Methods, 8, 441–
son, B.: Effective Visual Communication for the Quantitative 441, https://doi.org/10.1038/nmeth.1618, 2011b.
Scientist, CPT Pharmacometrics Syst. Pharmacol., 8, 705–719, Wong, B.: Points of view: Points of review (part 2), Nat. Methods,
https://doi.org/10.1002/psp4.12455, 2019. 8, 189–189, https://doi.org/10.1038/nmeth0311-189, 2011c.
Vanderkam, D., Allaire, J., Owen, J., Gromer, D., and Thieurmel, World Health Organization: World report on vision, 2019.
B.: dygraphs: Interface to “Dygraphs” Interactive Time Se- Zeileis, A., Fisher, J. C., Hornik, K., Ihaka, R., McWhite, C. D.,
ries Charting Library, available at: https://CRAN.R-project.org/ Murrell, P., Stauffer, R., and Wilke, C. O.: colorspace: A Tool-
package=dygraphs (last access: 2 August 2021), R package ver- box for Manipulating and Assessing Colors and Palettes, J. Stat.
sion 1.1.1.6, 2018. Softw., 92, https://doi.org/10.18637/jss.v096.i01, 2020.

https://doi.org/10.5194/hess-25-4549-2021 Hydrol. Earth Syst. Sci., 25, 4549–4565, 2021