Exploratory data analysis of Burkina Faso’s Raw daily meteorological records (1981–2021): data quality insights and potential outlier screening
Open accessThe DOI matches its Crossref record; 39 of the 39 references it cites were found in a registry or an index; 5 of the 8 sources its claims cite were checked for support.
Abstract
This study performs a national-scale quality screening of raw daily meteorological observations from Burkina Faso’s National Meteorological Agency, covering 10 synoptic stations over 1981–2021.
Overview
CoChat Summary
TL;DR
This study performs a national-scale exploratory data quality screening of Burkina Faso’s raw daily meteorological records from 10 synoptic stations over 1981–2021. It documents substantial missingness, wind-direction artefacts, coherent warming/wetting signals, and candidate inhomogeneities, while cautioning that trends are from non-homogenized series.
Key findings
- Overall missingness in the ANAM daily database is 10.41%; GRAD alone accounts for 64.25% missing data, and excluding GRAD the overall missing proportion drops to 4.65%. Bogandé has 36.7% missingness, Dédougou 10.7%, and Ouagadougou 0.3%; RAIN appears complete, raising concern about confusion between missing values and recorded zero precipitation.
- Wind direction shows suspicious artefacts: physically implausible values exceeding 360° were found (426° at Dori on 2020-05-13; 372° at Boromo on 2020-04-19), and WD-MEAN is discretized/capped at 288° for eight stations, with ECDFs terminating around 288° rather than 360°.
- Temperature anomalies show a coherent warming signal, with positive trends typically ~0.1–0.4 °C dec⁻¹ for T-MEAN and stronger T-MIN rates (≥0.3–0.6 °C dec⁻¹ at several stations); precipitation slopes are positive at all stations (+9 to +81 mm decade⁻¹), with post-2010 positive anomalies locally reaching ~+200 mm yr⁻¹.
- Wind-direction trends indicate veering at nine of ten stations, from ~+1.0° dec⁻¹ at Dori to ~+19.5° dec⁻¹ at Pô; Boromo is the main exception with backing (~−4.3° dec⁻¹). WS trends are mixed (e.g., Bogandé ~+0.84 m s⁻¹ dec⁻¹; Boromo ~−0.31 m s⁻¹ dec⁻¹; Dori ~−0.29 m s⁻¹ dec⁻¹).
- A preliminary homogeneity assessment found 54 of 81 station-variable combinations (67%) with significant breakpoints at p<0.05, clustered around 1997–2002 for temperature and 1990–1997 for RH-MIN/WS-MEAN; no correction was applied.
Methods
- Data: raw daily meteorological records from 10 synoptic stations operated by Burkina Faso’s National Meteorological Agency (ANAM), covering 1 January 1981 to 31 December 2021; no temporal aggregation or averaging was performed by the authors. Variables include RAIN, PET, T-MIN, T-MEAN, T-MAX, RH-MIN, RH-MEAN, RH-MAX, WS-MEAN, WD-MEAN, SUN-DUR, and GRAD.
- EDA workflow: station-variable completeness profiling and missingness mapping; box-and-whisker plots; standardized anomalies relative to the 1991–2020 climate normal; station-wise variability and OLS linear trend estimation; cross-station seasonal/interannual pattern comparisons; dispersion diagrams and climate variability maps.
- Homogeneity screening: Pettitt test and standard normal homogeneity test via the climatol R package on annual series for nine variables across nine stations (Bogandé excluded); no correction applied. Trend/anomaly analyses used a completeness criterion of ≥270 valid daily observations per year (~74%), and GRAD was excluded because of high missingness.
Limitations
- The paper states that no imputation was applied, no full homogenization was conducted, and all long-term trends are derived from raw, non-homogenized series; detected breakpoints are candidate inhomogeneities only, and trend magnitudes—especially warming rates for T-MIN and T-MEAN—may be partially inflated by instrumental step changes and should not be used for quantitative climate-change attribution without prior homogenization.
- The paper states that applicability of the workflow to other settings has not been evaluated and would require independent validation; the exact origin of the 288° WD-MEAN cap cannot be resolved from archived records alone and requires dedicated investigation using station metadata, instrument specifications, coding conventions, and archival procedures.
- The text makes evident additional constraints: Bogandé analyses were constrained to post-1997 for many variables because of severe data gaps; station-to-station trend comparisons can be biased by incomplete early records; RAIN completeness may reflect misinterpretation between missing values and zero precipitation; and GRAD’s high missingness limits its use.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
4 Valid · 1 Not valid · 1 Not checked · 115 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“For instance, [5, 6] demonstrated that inconsistencies in data collection practices contribute to significant biases in climate datasets, affecting the accuracy of climate modelling and predictive analyses. These biases can propagate through climate models, reducing the reliability of future projections, especially in data-scarce regions.”
What the sources say
- Partly supportedcites [5] Venema, 2012, Clim. Past sourceExperimental — accuracy not yet measured
The Introduction says changes in observation practices can create artificial biases in climate trends and that the resulting records are used to validate climate models. It does not establish that these biases reduce the accuracy of climate modelling or predictive analyses. The cited sentence does not attach this source directly to its further claim about future projections in data-scarce regions, which the paper also does not address. (LLM, conf=0.85)
- Partly supportedcites [6] Dhawan, 2024, Heliyon sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) says climate-model outputs contain significant biases that require correction. The Observed dataset section (page unavailable) says measurement inaccuracies can cascade into hydrological modelling, and the Introduction notes sparse weather-station coverage in remote regions. The paper does not establish that inconsistent data-collection practices cause those dataset biases or that they specifically reduce the reliability of future projections in data-scarce regions. (LLM, conf=0.83)
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“an exceptional precipitation event was documented in Ouagadougou with 263 mm of rainfall on 1 September 2009 [37]”
What the sources say
- Partly supportedcites [37] Engel, 2017, J. Hydrometeorol. sourceExperimental — accuracy not yet measured
Section 4 'Dynamical analysis' (Case 1: Ouagadougou, Burkina Faso, 1 September 2009) confirms the event, place, date and exceptional character: the airport synoptic station reported the highest 24-h accumulation ever recorded there, exceeding the second-largest value by nearly 100 mm, and the Abstract and Section 6 repeat its exceptionality. However, the value our copy gives is 261.3 mm (for the 24 h ending 0600 UTC 2 September 2009, i.e. the 0600-0600 UTC gauge day), not 263 mm; no passage in our copy states 263 mm (searched: '263 mm Ouagadougou 1 September 2009 rainfall total', '261.3 mm highest 24-h accumulation'). The magnitude attributed to this source is therefore slightly overstated/inaccurate, while the rest of the claim is supported. (LLM, conf=0.78)
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“which may partly reflect genuine hydroclimatic shifts rather than purely instrumental artefacts, given the known post-1990 s rainfall recovery documented across the Sahel [35, 36].”
What the sources say
- Not checked — no full text of the sourcecites [35] Sanogo, 2015, Int. J. Climatol. source
- Partly supportedcites [36] Nicholson, 2013, Int. Sch. Res. Not. sourceExperimental — accuracy not yet measured
Section 3.1 (Temporal Trends) of Nicholson (2013) does document a post-1990s recovery: the comprehensive analyses (Nicholson, Ali and Lebel, Lebel and Ali, Mahé and Paturel) 'generally indicated conditions of improved rainfall in the late 1990s and a few following years,' and 'Nicholson [3] confirmed a recovery of wetter conditions in much of the Sahel during the period 1998 to 2003.' However, the paper explicitly qualifies the spatial generality the claim asserts ('across the Sahel'): 'the degree of improvement varied geographically,' and Lebel and Ali's gauge-only series showed recovery 'confined to the eastern-most regions' with 'the drought continued unabated in the west (10°W to 15°W)' and only ~10% recovery in the central Sahel for 1990-2007. The same section also records that whether the drought ended in the 1990s 'depends in part on the region, years, and even months evaluated,' and that L'Hôte et al. concluded the drought continued through the 1990s — so the recovery is real but contested and regionally uneven, not uniform. The 'instrumental artefacts vs. genuine shifts' framing is the citing authors' own inference and is not addressed by this source (though 3.1 does note the difficulty of obtaining gauge data post-1990 and that Nicholson's 1998-2003 conditions came from TRMM while earlier ones came from gauges). (LLM, conf=0.72)
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Out of 81 station-variable combinations tested, 54 (67%) exhibited a statistically significant breakpoint at the 5% level (figures 13 and 14; tables S1 - S4 in supplementary materials).”
- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The figure highlights frequent gaps prior to 1996, likely reflecting technical constraints and recurrent station outages before the automation of the synoptic network.”
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
53 more claims
Create a free account to see all 121 claims and to verify the 1 reference still open.
- Probably not a substantial claim. 13
- The quote only partly states this claim. 1
- It may state two separate claims as one. 52
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“ANAM) operates 162 meteorological stations, including 10 synoptic, 12 climatological, 20 agrometeorological, and 130 rainfall stations”
Reported 162 · recomputed 172
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 3.4. Long-term trends and data consistency assessment
“To conduct this analysis, we combine annual variability curves with trend lines and climatic anomaly representations for the 1990-2021 period.”
§ 3.4.3. Precipitation (RAIN) and PET
“Figure 7. Annual anomalies in daily maximum (T-MAX), mean (T-MEAN) and minimum (T-MIN) air temperature at Bogandé, Ouagadougou, Bobo-Dioulasso and Ouahigouya for 1981-2021.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 3.3. Central tendencies and initial assessment of erroneous or outlier data
“a monthly scale provides a more robust balance between variability and interpretability.”
§ 3.3. Central tendencies and initial assessment of erroneous or outlier data
“short windows such as one week or one month are not adequate for reliable outlier screening. At least one full year of daily data is therefore needed”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 3.4. Long-term trends and data consistency assessment
“a given year was considered in the analysis only if the number of available daily observations exceeded 270 days”
§ 3.4.3. Precipitation (RAIN) and PET
“shown only for years meeting the data-completeness criterion ( ⩾ 270 valid daily observations per year for each variable).”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“When considering the complete ANAM daily-resolution database over the 1981-2021 period, the overall proportion of missing values is 10.41%.”
Reported 10.41 · recomputed 9.43
Citations
Cites
(IPCC), I. P. O. C. C. (2023) · doi:10.1017/9781009157896 · Cited by 6,972
A Non-Parametric Approach to the Change-Point Problem
Pettitt, A. N. (1979) · Applied Statistics · doi:10.2307/2346729 · Cited by 4,114
A homogeneity test applied to precipitation data.
Alexandersson, H. (1986) · Journal of Climatology · doi:10.1002/joc.3370060607 · Cited by 1,858
A Review on Outlier/Anomaly Detection in Time Series Data
Blázquez-García, A., Conde, A., Mori, U., & Lozano, J. A. (2022) · ACM Computing Surveys · doi:10.1145/3444690 · Cited by 1,102
Nicholson, S. E. (2013) · ISRN Meteorology · doi:10.1155/2013/453521 · Cited by 742
Benchmarking homogenization algorithms for monthly data
Venema, V. K. C., Mestre, O., Aguilar, E., Auer, I., Guijarro, J. A., Domonkos, P., Vertacnik, G., Szentimrey, T., Stepanek, P., Zahradnicek, P., et al. (2012) · Climate of the Past · doi:10.5194/cp-8-89-2012 · Cited by 368
Evaluation and comparison of satellite-based rainfall products in Burkina Faso, West Africa
Dembélé, M., & Zwart, S. J. (2016) · International Journal of Remote Sensing · doi:10.1080/01431161.2016.1207258 · Cited by 318
The U.S. Historical Climatology Network Monthly Temperature Data, Version 2
Menne, M. J., Williams, C. N., & Vose, R. S. (2009) · Bulletin of the American Meteorological Society · doi:10.1175/2008bams2613.1 · Cited by 300
Spatio‐temporal characteristics of the recent rainfall recovery in West Africa
Sanogo, S., Fink, A. H., Omotosho, J. A., Ba, A., Redl, R., & Ermert, V. (2015) · International Journal of Climatology · doi:10.1002/joc.4309 · Cited by 241
Martin, A., Weissmann, M., Reitebuch, O., Rennie, M., Geiß, A., & Cress, A. (2021) · Atmospheric Measurement Techniques · doi:10.5194/amt-14-2167-2021 · Cited by 155
Research works printed but not linked: 1
- Wahyono, 2020, J. Comput., 31, 42, 10.3966/199115992020083104004 · doi:10.3966/199115992020083104004
Cited by
Compaoré, C., Kiébré, R., Kébré, M. B., et al. (2026) · Atmospheric and Climate Sciences · doi:10.4236/acs.2026.163032
Self-citation
Metadata
- Authors
- Zamantakonè Ki, G., Bawindsom Kébré, M., Somda, W., Gandema, S., & Dabilgou, F.
- Affiliations
- Université Daniel Ouezzin Coulibaly (BF)
- Université Joseph Ki-Zerbo (BF)
- Corresponding author
- Somda, Wenceslas
- Dabilgou, François
- ORCID
- Venue
- Environmental Research Communications
- Publisher
- IOP Publishing
- Published
- 2026-05-01
- Volume
- 8
- Issue
- 5
- Pages
- 051012
- ISSN
- 2515-7620 (ISSN-L)
- OpenAlex
- W7160417203
- Cited by
- 1
- References
- 32
- Citations by year
- 2026: 1
- FWCI
- 3.03
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0/ (vor, 12 d)
- iopscience.iop.org/info/page/text-and-data-mining (tdm, 12 d)
- publishingsupport.iopscience.iop.org/questions/alternative-author-rights-policies/ (vor, 12 d)
- Locations
- Environmental Research Communications (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Environmental Science › Global and Planetary Change › Climate variability and models
- Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Precipitation Measurement and Analysis
- Physical Sciences › Environmental Science › Management, Monitoring, Policy and Law › Environmental Monitoring and Data Management
- Keywords
- meteorological data quality, data quality screening, outlier detection, missing data, homogeneity analysis, change point detection, climate normals, global radiation, long-term climate trends, data-scarce regions, Burkina Faso
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Exploratory data analysis of Burkina Faso’s Raw daily meteorological records (1981–2021): data quality insights and potential outlier screening. https://app.cochat.ai/papers/10.1088/2515-7620/ae68f0
- The paper
- Zamantakonè Ki, G., Bawindsom Kébré, M., Somda, W., Gandema, S., & Dabilgou, F. (2026). Exploratory data analysis of Burkina Faso’s Raw daily meteorological records (1981–2021): data quality insights and potential outlier screening. Environmental Research Communications. https://doi.org/10.1088/2515-7620/ae68f0
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.