The claim about "bombing for the last 8 years" is certainly true if one speaks of shelling. Aircraft operations have been negligible 2016-2021, or at least no source I looked at mentioned them. But shelling there was plenty, i.e. this was definitely not a "frozen conflict" even after 2015, unless one euphemistically applies that notion to the mostly static nature of the front lines during this period.
The OSCE SMM reports for Ukraine are probably the most detailed source when it comes to shelling, but they are somewhat unvently organized, making year-to-year comparisons difficult before 2018, at least when it comes to shelling. Their data on "explosions attributable to fire from MLRS, artillery, mortars and tanks" going backwards in time looks like this:
- 2021: 1,133 explosions
- 2020: 3,216 explosions
- 2019: 3,373 explosions
- 2018: 8,470 explosions
- 2017: I could not find separate data.
- 2016: no SMM-provided total that I could find, but they report (fig. 6 on p. 13) approximately 1,000-5,000 per month, so I'd guess 20,000-30,000 for the whole year. (Tank fire is seemingly not included in this graph, only MLRS, artillery, and mortars.)
Alas the location of the shelling is not numerically broken down by where it landed (Kyiv-controlled vs. LPR/DPR-areas) although these reports do have some maps for a general impression of where the shelling concentrations were... i.e. mostly on the contact line.
On the other hand, the majority of the civilian casualties caused by these occurred in the LPR/DPR-controlled areas, e.g. for 2016:
The majority of casualties – 281 out of 442 – occurred in non-government-controlled areas, with 57 civilians killed and 224 injured while 148 casualties (28 killed and 120 injured) were reported in government-controlled areas. Additionally, 13 cases (three killed and ten injured) occurred in locations not fully controlled by any of the sides.
The combined report for 2017-2020 points to a similar distribution of civilian casualties (dead and wounded), although somewhat more skewed than before.
- Government-controlled areas: 223 in Donetsk region and 47 in Luhansk region
- Non-government-controlled areas: 513 in Donetsk region and 144 in Luhansk region
- Areas not controlled by either side: 14 in Donetsk region and five in Luhansk region
It seems that most civilian casualties 2017-2020 occurred in "hotspots" that were contested by both sides, rather than by shelling in areas further to the rear (although there was some of that too judging by the maps):
Four hotspots that straddle both government- and non-government-controlled areas of the contact line, three in Donetsk and one in Luhansk region, accounted for nearly 75 percent of civilian casualties due to shelling and [small arms and light weapons]-fire.
In all those reports, the vast majority of civilian casualties were due to shelling or delayed explosions caused by unexploded ordnance or mines/IEDs etc., rather than lighter/direct weapons fire, even though the latter category constituted an overwhelming proportion of the overall "ceasefire violations" that the SMM tracked. (E.g. 312,544 violations in 2018; 93,902 violations in 2021.)
As noted in the other answer, during all this time, the contact line ran pretty close to the city of Donetsk, by far the largest city in this conflict area (pop. 900K and in DPR hands), and so [at least] its outskirts were hit by shelling judging by the 2021 maps, and likewise for the outskirts of Luhansk city (pop 300K). The outskirts of Horlivka, which is roughly the same size as Mariupol or Luhansk, were pretty much shelled year after year, as the contact line ran between Horlivka and Toretsk (pop 31K), which however was controlled by Kyiv, and was "only a few kilometers away from the separatists-controlled Horlivka" (as Wikipedia puts it). Krasnohorivka (pop. 16K) was in a similar position to the east of Donetsk city and Avdiivka (pop. 32K) to the north thereof was contested by both sides. In between Horlivka and Luhansk the contact line also ran though somewhat populated areas, like Zolote (pop. 13K), also contested by both sides. So, yeah, the contact line ran close or downright through some fairly populated areas.
Now 4-100 shells/explosions per day (depending which year you want to consider 2016-2021) definitely is not peace, but compared to 60,000 per day in 2022, if the Western estimates of the latter are correct... (there's no more OSCE on the ground to count every explosion etc.)
User gunterze has kindly went through the 2022 OSCE reports (before they withdrew) and added these tallies:
The increase at the end of Feb had been noticed by the international press, as well as by Russia. Ukraine and LPR/DRP trade accused each other of bearing responsibility for this increase.
(Putin announces the "necessary, long-overdue decision" to recognize the independence of LPR/DPR on Feb 21, and instructs his Ministry of Defense to send in Russian troops. The next evening Putin says he recognizes the LPR/DPR territorial claims over the entire Donbas region so they "could restart their fight to establish their territorial integrity", despite the fact what LPR/DOR occupy only about a third of that on Feb 21, and contra to what deputy FM Rudenko said earlier on Feb 22.)
