Trend Analysis of Temperature Data for Narayani River Basin, Nepal (Version 3, Revised)
|Reviewer 1 Chinh Ta Huu National Center of Hydro-meteorological Forecasting, Vietnam||Reviewer 2 amrit thapa International centre for integrated mountain development (icimod)||Reviewer 3 chinh ta huu National Center for Hydro-Meteorological Forecasting, Vietnam|
|Agreed, preparing report||Agreed, preparing report||Agreed, preparing report|
|Agreed, preparing report||Agreed, preparing report||Agreed, preparing report|
|Agreed, preparing report||Agreed, preparing report||Agreed, preparing report|
Chand, M.B.; Bhattarai, B.C.; Pradhananga, N.S.; Baral, P. Trend Analysis of Temperature Data for Narayani River Basin, Nepal. Sci 2019, 1, 49.
Chand MB, Bhattarai BC, Pradhananga NS, Baral P. Trend Analysis of Temperature Data for Narayani River Basin, Nepal. Sci. 2019; 1(2):49.Chicago/Turabian Style
Chand, Mohan Bahadur; Bhattarai, Bikas Chandra; Pradhananga, Niraj Shankar; Baral, Prashant. 2019. "Trend Analysis of Temperature Data for Narayani River Basin, Nepal." Sci 1, no. 2: 49.
Article Access Statistics
National Center of Hydro-meteorological Forecasting, Vietnam
International centre for integrated mountain development (icimod)
National Center for Hydro-Meteorological Forecasting, Vietnam
University of Padova
Chand et al. investigate the spatio-temporal temperature dynamics in Narayani river basin, located in Nepal. They analyze temperature data from six stations, distributed at different elevations within the basin, data availability ranging from 1960-2015. Since all the stations present significant data gaps, they use multiple regression and empirical mode decomposition to fill these gaps. The comparison between filled data and temperature datasets provided by different freely available sources shows a good correlation, supporting the reliability of the filling procedure. The data analysis shows a significant increasing temperature trend after 1970s in the majority of the stations (mean warming trend of 0,03°C per year), with higher rate of increase during monsoon season and winter. Furthermore, Chand et al. finds a relation between temperature and elevation, showing a temperature lapse rate with altitude of 0,006°C per meter within the investigated basin, which can be very useful considered the lack temperature data in the higher Himalayan region.
Overall I found the manuscript well written and it highlights once again a significant issue of our age, climate change and global warming, focusing on an area very sensitive to the global temperature increase like the Himalayan region. The analyses and interpretations are sound, but in my opinion the discussion about some points should be expanded. In my view, the manuscript is suitable for publication in Sci with a minor to moderate revision.
Here it follows a list of suggestions and observations the authors might want to consider:
- Data clearly show that, until 1970, there is a significant cooling trend and that this cooling trend last longer in time at many of the measuring stations. Can you explain this behavior? Can you provide any hypothesis regarding the cooling trend preceding the clear warming trend that follows?
- Figure 1: I would appreciate if you could highlight the Narayani River in the map.
- Paragraph 3.1: “data gaps more than 5 years” I think it would be better to write “data gaps longer than 5 years”.
- Paragraph 3.2: even if it is quite clear, I think you should briefly explain why the skewness of the data follows from the statistical analysis.
- Figure 2/3/5/6/8/9: please adjust the dimension of the figures in order to make them more homogeneous with the others figures and the rest of the manuscript.
- Figure 4: I think that all the boxes in figures 4 should be of the same dimension in order to make them comparable, also the last one.
- Paragraph 3.5: “Shrestha et al.  and .” It seems that a citation is missing.
“that is, the MK test is a non-parametric (distribution-free) test.” I think there is an error.
- Paragraph 4.1: both in Paragraph 4.1 and Figure 6 you show and describe all the components obtained with the EMD decomposition, but then you just discuss and focus on the last, and probably more interesting, one (IMF 6). I think that, since you show all the components of the decomposition, you should spend some words on each one of them, explaining the meaning of the results you obtained, at least briefly.
- Paragraph 4.2: as you highlight, most of the measuring stations show a trend break in between 1970 and 1979, while the measuring station of Langtang shows the same trend break in 1993. Can you explain that? Or does this just depend on the lack of data for the measuring station of Langtang before 1988? In both cases, I think you should discuss this point.
- Paragraph 4.3: “Higher monthly mean variance in April is the indicator for changing the number of days in winter and summer. Hanjra and Qureshi , KC and Ghimire  also observed more frequent warmer days and less frequent cooler nights throughout Nepal. Annual maximum temperature is in an increasing trend with hot summer days while minimum temperature is in decreasing trend with cool winter days  which is reflected by the higher variance in the month of April and December.” I think this is a very interesting point and the discussion should be expanded.
“Significant warming trends were observed at five out of six stations in the monsoon, winter and post-monsoon seasons and at four of six stations in pre-monsoon season (Table 4).” Can you suggest an explanation about that?
- Paragraph 4.4: “The discrepancies between different seasons are due to differences in relative humidity ) and incoming solar radiation .” I think it would be useful to explain how relative humidity and solar radiation affect temperature causing the highlighted difference between different seasons.
- Figure 12: I think that MAAT should not be displayed as the last point of the plot, but it should be shown by itself. This representation might be misleading. Please, add the axes labels both here and in Figure 13.
- Paragraph 4.5: I think that MAAT values should not be considered when fitting the trend lines shown in Figures 14-15-16-17. It won’t change that much but it seems uncorrect.
Response to Reviewer 4Sent on 09 Jul 2020 by Mohan Bahadur Chand, Bikas Chandra Bhattarai, Niraj Shankar Pradhananga, Prashant Baral
University of Alberta
Overall this manuscript is well prepared. The authors analyses air temperature data for six stations in Nepal. A gap-filling process was used to interpolate missing data. The gap-filled in situ data was compared with derived products (modeled data, remotely-sensed data).
Improved language clarity would greatly enhance the manuscript.
The authors do a good job of including relevant studies of similar data over a similar. Better distinction between this study and previous studies would help the reader understand the specific contribution of this piece of work.
Introduction – very good review of existing work but I found I got lost in the details. Instead of listing each study separately is it possible to synthesize the results more generally? Or perhaps include a paragraph break to more clearly demonstrate summary of in-situ vs model-based studies?
3.3 Gap filling
Please explain or elaborate on gap filling process and use of stations. In Section 3.1 you describe how stations were selected that represent different physiographic zones and climate ranges. Given this statement how would these differences affect the gap-filling process? How are these differences addressed in the gap-filling process, if at all? What would be the impact on the ‘filled’ data?
3.4-Trend break detection
Longest record used for the EMD analysis. This makes sense but you need to state if and how the results for that station might be applicable to the other stations in your region. You specified that your stations cover different physiographic and elevation zones. Are the results obtained in the EMD analysis more a reflection of the broader climate and less so of the micro/local climate? If this is the case then you could assume what is happening at Nuwakot is probably also happening at the other stations. This is just something to consider.
Was any attention paid to autocorrelation when conducting your trend analysis? If so how? If not, perhaps something to consider.
MODIS – include additional references, see:
Wan Z 2013Collection 6. MODIS Land Surface Temperature Products. Users’ Guide. [and references therein]
Given the climate of the region, the absence of data when clouds are present is might be an important factor affecting the usefulness of optical satellite data. MODIS data used here are averages of all available cloud-free scenes within an 8-day period. MOD11A2 values are produced when there is a minimum of one cloud-free scene within a given 8 day period. This means that variability within that 8 day period can introduce variability in the MODIS record that is not physically real. In areas affected by cloud cover this could be a significant issue. The number of days used for each retrieval is provided in the QA layers.
3.6 – please include location and date of data access of data products either here or in the references.
Add reference to MODIS data product (example is from LP DAAC):
Wan, Z., S. Hook, G. Hulley. MOD11A2 MODIS/Terra Land Surface Temperature/Emissivity 8-Day L3 Global 1km SIN Grid V006. 2015, distributed by NASA EOSDIS Land Processes DAAC, https://doi.org/10.5067/MODIS/MOD11A2.006
4.4 - Temperature gradients
The analysis of ‘lapse rate’ isn’t too convincing as the stations are in different physiographic regions. What is really being calculated is the line between temperature and elevation from your sample of stations. Lapse rate is usually at a given location. This is probably more a matter of terminology? However, in looking at the plot (figure 11) four of the stations are around 1000 m a.s.l. so it is clustered to the left. What I find more interesting than the lapse rate is that the seasonal difference in temperature is quite small at the high elevation station compared to the other 5 (I can’t tell if the mid-elevation station has a larger spread than the low elevation stations). There is something interesting here but I don’t know if it comes across the way it is described.
4.5 – Comparison of station data with freely available global climate datasets
Data availability did not permit comparison with in-situ data and derived data over a common time period. Please state this and tell the reader more clearly how you compared the datasets. i.e. for each derived data product, in situ data for the corresponding period of record was compared. I am curious as to why only 2015 was used for MOD11A2. Data for this product are available beginning in 2000.
Given that you are comparing different time periods, and your trend analysis showed changes in temperature, would you expect to have similar average temperature values? Please provide some mention of this caveat to ensure your comparisons are properly interpreted by the reader.
Line numbers refer to lines of text on each page. Figures, Tables and captions not included in numbering.
Overall, suggest yr-1 instead of per year, and decade-1 instead of per decade
P1 – Abstract
L2: downstream of what?
L3: glacier ice
L5: Near surface (2m) temperature data? Specify somewhere near the beginning that you are looking at air temperature, except of course where you are looking at skin temperature.
L6: m a.s.l.
L8: remove ‘the’ from ‘the missing data’
L9: add ‘a’ to ‘a Mann-Kendall’
L9: revise ‘for testing the’ to ‘to test the’
L10: ‘trend’ should be ‘trends’ plural
L12: by ‘mean increasing trend’ do you mean the mean of all six stations? Please clarify
L3: remove ‘the’ from ‘the global’
L4: ‘are significantly rising since last decade’. Suggest revising to: ‘have been rising significantly over the last decade’
L6: change by 0.12°C per decade to ‘to 0.12 decade-1’
L7: change ‘the rate of increase of temperature are different’ to ‘the rate of temperature increase is’
L8-9: note – mountain glacier melt contributes to the eustatic component of SLR. [as opposed to thermal expansion]
L1: revise ‘This warming are’ to ‘this warming is’
L6-7: you mention ‘several studies’ but only once citation. Does this citation discuss multiple studies or is more than one reference necessary. COLLEEN TO CHECK
L17: revise ‘by 0.03°C’ to ‘of 0.03°C
L18: and ‘by 0.38°c’ to ‘of 0.38°C
L19: suggest revising to something along the lines of: ‘found that increases in maximum temperature (0.05 °C yr-1) were larger than those of the minimum temperature… with the largest change observed during the pre-monsoon season…’
L22: revise ‘in average temperature at a rate of’ to ‘in average temperature of’
L24: revise ‘The model’ to ‘Model’
L24: ‘the Koshi River basin’ (add ‘the’)
L25: suggest revising ‘due to rise in temperature’ to ‘due to increasing temperatures’
L27: ‘the Koshi trans-boundary…’
L45 (study area): revise ‘between latitude….85.83°E’ to ‘between 25.49°N-29.28°N and 85.02°E-85.83°E’
L47: ‘at the southern edge’
L48: km² instead of sq. km
L48-49: ‘It encompasses a total area of…, and area within Nepal is about…’ to ‘It encompasses a total area of 46,300 km², 32,104 km² of which is in Nepal’
L3: ‘The upper part of this basin is covered with snow’ (add ‘is’)
L7-8: suggest revising to ‘large variations in climate over the basin’
L10: suggest revising to ‘in the monsoon season’
L15 (methods, data availability): specify/clarify if it is daily air temperature.
L16 (methods, data availability): ‘were selected (Table 1) from different …’ to ‘were selected (Table 1), each representing different…’
L19: ‘and have data gaps longer than 5 years’. Unclear if you mean none of the stations have data gaps >5yrs or if they all have data gaps longer than 5yrs.
L19: suggest revising to: ‘Langtang station has the shortest temperature record’
L21: are not uniform in what sense? Slightly ambiguous.
L2: suggest: ‘Langtang (Station ID 1000) is the highest elevation station among the six stations but has the fewest number of years of temperature observation…’
L2: suggest revising ‘All the available’ to ‘All available’
L3: suggest: ‘Outliers were identified and removed…’
L5: revise to ‘assumption were applied’
L7: stations (should be plural)
L8-14: Did you use ‘Statgraphics’ to do this procedure? This section is a bit unclear.
L9: suggest revising ‘were slightly less than’ to ;was slightly lower than’
L1: either method (singular) and ‘is’ (singular) or ‘methods’ and ‘are’
L7-9: lacking a bit of clarity. Are all stations used in the rest of your analysis but in this case (for the change-point analysis) you are using a single station and you chose the one with the longest record? Are results from this one station then applied to the other 5 stations? How did you use the information from this analysis to inform the rest of your analysis? i.e. did you then compute trends before and after the identified change-points?
L13: ‘the Mann-Kendall’
L15-16: a bit unclear to me. How were these breaks identified at most of the stations?
L22: suggest revising ‘should be’ to ‘are’
L23: suggest revising ‘associated to’ to ‘associated with’
L18:do you mean NOAA (instead of Noah?)
L19: ‘were used’
L20-30: references required. See general comments.
L26-27: unclear – Monthly average… were further averaged to obtain monthly averages’ Do you that you took an average of all stations? Please clarify.
L37-47: language could be more concise and direct. This would likely address general comment concerning time periods and comparisons.
L43-45: why only 2015? (see general comments)
L3: ‘EMD methods were used to detect temperature trends at six’
L15: ‘mean annual temperature anomalies’
L17: unclear – ‘few extreme rise and fall are recorded’ Do you mean that there were few anomalously high or low temperatures?
L27: How does your study differed from Shrestha et al? Same or different stations? Same or different time period?
L30: at the 96%
L31: ‘that starts in’
P11 Table 3 – maybe period 1 and period 2 because it’s not really the first half and second half?
L2: ‘significant’ – do you mean ‘statistically significant?’
L1-2: are those values the average of all 6 stations? A bit unclear.
L2-3: language a bit unclear. Suggest revision along the lines of: ‘For the period 1970-2015 all stations except Chapkot had statistically significant trends’ (or something similar; more clear language in remainder of paragraph might also help)
L8: why all of a sudden is 1993-2015 being used? Where did this come from. Please clarify.
L11-13: these two sentences seem to contradict each other. Or maybe I am missing something?
L6: excellent to compare your results with other studies and to place your results in this context but sometimes it is unclear if or how your study is different from these existing studies. If your work is unique and adds something special I want to know that.
L10 – Heading – perhaps monthly and seasonal trends?
L11-12: suggested revision ‘…at each station (Figure 8) is lowest in January and highest during June-August.’
L13: perhaps say ‘the lowest’ instead of ‘minimum’ because using minimum can make a reader think you computed the average, maximum, and minimum. Similar confusion occurs elsewhere [e.g. next line with ‘maximum’]
L18: should be ‘months of April and December’ (months plural)
L25: suggest ‘The calendar year’ instead of ‘A year’
L26-27: ‘to assess the seasonal temperature trend’
L33: ass commas – ‘per year, respectively, and those’
L33: statistically significant at what confidence level?
L34-35: ‘at the basin and regional scales’
L5: ‘The lapse rates were calculated’ (lapse rates plural)
L6: suggest revising to: ‘The largest (…) and least negative temperature gradient…’.
L7-8: suggest: ‘…whereas … had the same lapse rate (…)’
L12: perhaps ‘The more gradual’ instead of ‘the very less steep’
L13: ‘are due to the strong’
L18 – an introductory sentence to this section (section 4.5) would be helpful.
L18-19: suggest revision to something along the lines of: ‘There is little difference in the mean monthly temperatures from the modeled temperature datasets (GLDAS 2.0, CHELSA 1.2) at…’
L20-21: A bit confusing here between time periods and datasets. This relates to general comment about this comparison. It’s not clear if you compared WorldClim data with in situ data for two separate time periods (1970-2000, 1979-2013) or if one time period was used for one product and another time period for another product.
L22: suggest ‘is expected’ or ‘is not surprising’ instead of ‘is obvious’
L29: It is the mean monthly LST temperature (not the mean monthly temperature from LST) MODIS is giving you skin temperature, your in-situ data is air temperature.
L1-14: series of disparate sentences should be combined into a single paragraph.
L1-14: what is meant by ‘satisfactory correlation’?
L14: This line is important. Perhaps a few words to elaborate on the importance of this statement?
L7: suggest ‘to identify’ instead of ‘to find out’
L8: what is meant by an ‘acceptable’ correlation?
L18: is this the average of all 6 stations?
L21-22: see general comment about lapse rate.
L23: although I don’t disagree here I am not sure where this conclusion came from in relation to the results presented. Perhaps it fits better as a general comment or statement? Perhaps I missed something?